Publishers enter 2026 facing unrelenting pressure to innovate with generative AI, colliding with the need to protect editorial standards and audience trust. As AI slop floods the media ecosystem, publishers are realizing that their competitive advantage isn’t how fast they can use AI, but how safely they can integrate it, because it impacts trust and the direct relationships they have with their audiences.
Recent research in Canadian newsrooms reveals a cautious reality. Amid the hype, media leaders are not hiring for AI expertise. Instead they’re doubling down on core journalism skills, treating AI as an efficiency tool rather than a replacement for human judgement. It is important to note that this research focused specifically on the editorial use of AI, where the priority is the preservation of audience trust, rather than on use of AI for business growth or commercial purposes.
Terra Tailleur and I conducted interviews with CEOs and editors-in-chief at 12 media organizations, ranging from public broadcasters, national news outlets and wire services, to regional dailies and independent digital startups, over the course of eight months to explore how Canadian newsrooms are using and adopting AI in their editorial practices. Three practical approaches to responsible AI adoption have emerged for media leaders:
Trust-first guardrails
Our findings suggest a growing divide based on the size of outlets. Larger ones, like The Canadian Broadcasting Corporation (CBC), The Globe and Mail, and the Canadian Press have robust guardrails and policies. Smaller outlets, constrained by time and resources, often rely on informal oversight. This relegates ethical boundaries to individual intuition rather than documented standards.
In a recent HEC Montreal study of over 400 journalists, 36% of those surveyed were unaware if their organization even had an AI policy. Thus, publishers face an operational challenge not in drafting policies, but in driving them clearly and consistently from the executive corner office to reporters’ desks.
Small newsrooms don’t have the budgets of national broadcasters or wire services. So, they are forced to create simpler, more practical models. At Cabin Radio in Yellowknife, for example, editor Ollie Williams says that AI experimentation “happens so far off the side of the desk that it’s like the movie Inception and it’s like the desk has folded back in on itself three times before I get to it.” Therefore, he doesn’t have time to research AI uses and meet with staff to develop formal policy because he’s too busy running day-to-day operations with his editorial team of four.
For resource-strapped newsrooms, a simple governance model can start with a clear approval process, requiring editor sign-off for all AI use. Newsrooms should actively disclose when and how they use AI, prioritize transparency with audiences, and train staff on verification rather than technical skills.
For 2026, the goal for smaller publishers isn’t to draft a 50-page manual. Rather, they should try to establish living documents that provide clear guidance for daily tasks. As DCN outlined, “staff need clarity about when and how to use AI and how to validate its output.”
Upskilling and in-house training
The media organizations we spoke with weren’t hiring engineers with a surface-level interest in news for editorial purposes. Instead, they are conducting in-house training to fill the tech gap. The focus has been on upskilling existing staff who already understand the brand’s voice, ethics and audiences is more effective than bringing in tech-first hires who may lack journalistic DNA.
Tools can be taught in a controlled environment. At the CBC, for example, they aimed to train every employee, from summer hires to 30‑year veterans, with a full‑day AI program adapted from Radio‑Canada. This approach keeps AI adoption grounded in newsroom culture, not vendor experimentation.
Priority on core journalism skills
Across the board, editors-in-chief emphasized that AI expertise comes second to strong reporting fundamentals. More than the ability to write a clever prompt, they were looking for curiosity, critical thinking, strong judgment. They also focused on an ability to interview people, build sources, and find good stories. All of which are fundamental skills that define quality journalism.
Toronto Star editor-in-chief Nicole MacIntyre told us that while the next generation will, “absolutely need to embrace the tools that can help them work smarter and more efficiently… their success will still depend on having the core skills to adapt in a fast-changing media environment: curiosity, critical thinking, strong judgment, and a commitment to truth.” Ultimately, AI fluency matters, but only on top of reporting fundamentals.
Considerations before rolling out ambitious AI products
The test for managers and boards this year lies in day-to-day governance: Are newsrooms giving editors clear boundaries on generative AI, or leaving it to gut instinct? Before scaling AI deeper into editorial routines, here are pointed questions to ask:
Do frontline editors know the exact off-limits line for generative AI, or are we relying on vibes?
Are we allocating time and tools to verify AI-assisted content, or prioritizing speed over trust?
Do budgets support upskilling current staff on AI literacy, or are we waiting for perfect new hires?
With a third of the industry unaware of AI policies, what steps turn intranet PDFs into daily habits?
Publishers entering 2026 face practical choices about how AI fits into editorial workflows. The Canadian newsrooms we spoke with are moving cautiously, focusing on guardrails, staff training, and core reporting skills rather than rapid experimentation. Their approach suggests that AI in journalism will be shaped less by hype than by the daily realities of newsroom capacity, oversight, and editorial judgment.
A decade ago, the concept of liquid content emerged as a response to the fragmentation of devices, platforms, and audience consumption habits. The idea was that publishers could create modular or atomic stories that could be versioned cheaply for multiple channels, increasing reach.
Except that it did increase workload, and the work was too manual for publishers to sustain. But the idea may have just been a decade ahead of its time.
Liquid content—sometimes called structured content, modular content, or atomic content—is the practice of breaking stories into reusable components that can be reassembled for different formats, audiences, and platforms. The concept first gained traction a decade ago as publishers struggled to meet rising distribution demands across web, mobile, and social channels. Back then, the tools weren’t mature: versioning content for every platform created more work, not less. Today, generative AI, improved personalization, and cross-platform automation are reviving the strategy. With AI now able to help structure, maintain, and re-render content, publishers are revisiting liquid content not only to serve current channels more efficiently, but to future-proof their operations for distribution platforms and channels that don’t yet exist.
For David Cohn, senior director, generative AI innovation for content and the newsroom at Advance Digital, the concept itself isn’t new, just newly practical. From 2021-2014, he ran Circa , which was a mobile news app that restructured news into atomic units like events, statistics, quotes, and images that could be resurfaced and reused. The problem was scale: even with a globally distributed team working 24/7, Circa could barely keep up with the biggest stories of the day because every atom of news had to be created by humans. It wasn’t sustainable.
“Now, we just need more compute,” Cohn says. “And, while compute is expensive, it’s a lot cheaper and a lot more doable.” So, he says, it’s time for media execs to revisit “those ideas around structured content and see if they are more attainable and more worthy of our attention.”
The data layer: from story to query
Cohn’s current work at Advance Digital, while still in the thinking and testing phase, treats liquid content less as format differentiations, and more like infrastructure. Instead of a one‑off story, an article is framed as a query across a structured datastore of atomic units like facts, quotes, statistics, dates, and other fields extracted from reporting.
An article then becomes more like a SQL query that joins pieces of verified information and uses an AI layer to render them as a traditional story, a video script, or another format altogether. In other words: it is rendered whatever way people prefer to have their content delivered .
“But, what we’re in charge of is, I’ll use the word purity, of the data: how well it’s structured, how well it’s maintained, how much we add to it,” Cohn says. “That can be a proprietary data set. It is in our ability to create those proprietary data sets that we all of a sudden have value over large, giant technology corporations.”
Cohn is quick to flag that structured, liquid systems excel when audiences want clarity and up-to-date facts but are a poor fit for narrative non-fiction like the classic Gay Talese’s Frank Sinatra Has a Cold. “This type of structured content would be terrible to apply to that story,” he explains. “It’s about knowing what type of news and information this is great for and what types it is not. There are certain types of stories that are best told as a narrative. It’s challenge to make sure it’s applied in the right way.”
But, the upside, according to Cohn, is creating a massive system of highly adaptable content. “Once you start to do this at scale, you end up where the parts are more valuable,” he says. “There’s a number of ways that we can imagine that that kind of information becomes useful once it becomes a database that you can query.”
Digital media companies’ vast archives may become useful datasets in this framework. At SXSW this year, Lee Enterprises CIO Virginia Fletcher suggested that media outlets get their archives in order, wrote Alex Mahadevan for Poynter. With a structured database of articles, media companies can convert what they have into different mediums.
This sentiment was echoed by Jane Barrett, head of Reuters AI strategy, this year at the Nordic AI in Media Summit, noting their vast archive has inconsistent metadata. “Generative AI gives us the possibility to classify and structure our archive more effectively,” she said in a panel discussion. This moment is pushing us to think about our content as data. As Barrett put it, “it’s kind of giving us a good challenge to think now about our content as data and what needs to be true to turn our content into data that then can flow into these liquid experiences.”
The knowledge layer: encoding journalistic value
While Cohn’s focus on liquid content is its data system, researcher Sannuta Raghu, who heads Scroll’s AI Lab, pushes the idea one layer deeper, to what she calls the knowledge layer of journalism.
During her ICFJ Knight Fellowship in 2024, she mapped the structural layer of journalism, and compiled her findings in what she calls the Directory of Liquid Content. The directory is a scalable and modular taxonomy designed to map, describe and standardize how digital news is structured, styled and surfaced.
“Liquid content is not necessarily moving something from one format to another,” she says. “It’s moving something, or it’s using information such that it can be contained in any container on demand.”
Liquid content can move between various containers – from an article to a podcast, a calculator, or a decision tree, she says. Within these, it can also be formatted in various styles, as an inverted pyramid, an interview or an explainer for short TikTok videos.
“So, it’s making sure that you are able to have information that is dynamic enough that can be adapted to various containers, with high fidelity to the source where it came from,” she says.
Months into that fellowship, she was granted another Fellowship at the Reuters Institute, where she developed the News Atom to codify journalism’s epistemic layer. Raghu explains when AI parses a news article, “what they’re essentially doing is they’re looking for statements which can be converted into causal relations and subject object predicate,” she says.
When a news article is used as raw material to answer questions, AI will remove the intentionality of the signals a journalist has added. These can include the temporal aspect, words that describe when something happened, complex quotes and nested quotes.
“There are many subtle examples,” Raghu says. “Words like likely, reportedly, allegedly, very intuitive to us as journalists, but are considered extra when it comes to form.”
The News Atom provides a way for the news industry to structure, define and embed its own meaning rather than having that meaning imposed by external systems. Raghu believes if publishers don’t define and encode their own values, like trust, attribution, temporal precision, into the data they produce, those values won’t exist for AI models.
Liquid content in 2025 and beyond, then, is about survival in an ecosystem and consumption experience of on-demand form. This on-demand form transformation is already happening around the world, Raghu points out. Take for example, Google’s NotebookLM, which converts PDFs into mindmaps and podcasts.
“For us, it’s about making sure that you are codifying some of the practices that are intuitive to us. Like the inverted pyramid, for example, is a very journalism format. But, it needs to be codified and taught to a model in a very deterministic way so that it can be replicated and converted into multiple formats.”
For publishers, the question is whether to codify their practices and values into these systems, or whether those definitions will be written by others. As Cohn puts it: “We’re in a world where generative AI feels like the beginnings of the internet again, where we can fundamentally rethink assumptions.”
“We want to both understand it and have a solid foothold in it, rather than just be steamrolled by it,” Cohn says. “Has this fundamentally changed everything for Advance? No, but what we’re doing is to prepare if that does become the case, it’s on our terms.”
The shift toward liquid content is not simply a tactical experiment. As AI systems increasingly generate fast, unverified content in every format, publishers’ defensible advantage is the integrity and structure of their own data. Pursuing liquid content at scale may require rethinking stories as more than finished containers and embedding core journalistic values — attribution, verification, temporality, precision — into the underlying layers where content is created, stored, and reused.
This approach also raises the stakes for content architecture. Unstructured archives function as low-value files that can be scraped or summarized by AI. Structured, verified, and queryable archives become high-value assets, built from components that cannot be easily replicated with the same fidelity. Liquid content, in this context, is not a mandate but a strategic shift: a way for publishers to strengthen the foundations of their work and ensure it can adapt to the formats and distribution environments still to come.
Publishers are rethinking how they show up on social platforms in 2025, places where Gen Z and Gen Y increasingly discover content through algorithms. Recent DCN research reveals that these generational audiences’ media diets are dominated by algorithmically surfaced video. Individual creators now outperform traditional media brands on trust, creativity, and entertainment value. This trend challenges publishers to adapt their approaches while maintaining editorial integrity.
The visibility challenge is fundamental: algorithmic feeds surface content based on engagement rather than quality or brand recognition. Publishers are no longer guaranteed an audience simply because someone follows them. Every post competes equally with creators, memes, and friends’ updates, making traditional media brands just another scroll in the feed.
Media leaders from The Guardian, The Verge, Eater, share how they’re maintaining relationships with younger audiences through platform-native content, personality and experimentation.
The creator connection
Andrew Hare, senior vice president of quantitative research at Magid, and co-author of DCN’s report, “Decoding video content engagement: Gen Z and Gen Y in focus,” points out that creators have fundamentally reset audience expectations. “Creators rewrote the playbook: they win on originality, honesty, and personality,” he says.
Hare notes that the definition of quality has shifted. “At one point it was all about the image, professional appearance; it was basically something you would see on television. Now we realize that when we move into these digital spaces, other factors impact trust. Do you believe the creator or brand? Is it an authentic message? Is it real and unfiltered to some degree?”
The nature of engagement with Gen Z and Gen Y is both reactive and participatory. “I think the remixing, commenting and sharing nature of these platforms has been something that brands have been a bit slower to embrace because it means the conversation might be evolving, and they cede some control,” Hare says. “But, as we are seeing, it is this living conversation that creators have innovated upon and the very thing that provides value on social and YouTube.”
Gen Z trust friends and family most at 88%, followed by individual creators at 79%, with brands at only 61%. Hare says younger generations are suspicious of brands when they see them in their digital spaces. “Brands need to build trust in much the same way creators have in order to prove they can be authentic and real and provide relevant value. Otherwise, they are just noise and in the way of more engaging content.”
Kate Scott-Dawkins, global president of business intelligence at WPP Media, notes this aligns with broader shifts. “We have also seen this shift in terms of ad revenue globally, a majority going to creator platforms and UGC platforms rather than legacy media.”
Tuning tone: platform fluency and authenticity
Publishers emphasized they’re not lowering editorial standards to compete on social platforms, Rather, they’re translating their journalism into formats that work in algorithmically-driven spaces. The challenge they outlined was capturing attention in those critical first seconds while staying true to their standards.
In a world where audiences connect with individuals over brands, developing recognizable voices within their own ranks isn’t just a creative choice, it’s a strategic necessity for publishers. Journalists who embody the brand’s values can humanize it, build trust, and compete with the authenticity that individual creators command.
Success, then, depends on how effectively newsrooms and social teams collaborate, from story development to publication.
At The Guardian, Max Benwell, deputy head of audience for The Guardian US, says encountering content off-platform doesn’t preclude building direct relationships. “Because people are encountering our content off platform and maybe without context about the Guardian, it doesn’t mean they can’t build a direct relationship with us over time,” he says. “But the key is all about looking at the data we have and making sure our off-platform content is as engaging as possible. Then, because of how the algorithms work, we can trust that they’ll start seeing more and more of us.”
Benwell identifies short-form video opportunities during newsroom meetings and reaches out to reporters. “We’re also trying to create a mutual relationship where they, when they’re working on a story, can come to us and flag it way ahead of time.”
Flagging stories early means social teams can plan authentic, substantive videos rather than scrambling to repackage finished articles, creating content that competes with creators on engagement while maintaining journalistic standards.
Guardian social content is held to identical standards as the website, from brief to script to final edit. If you’re hearing from a Guardian journalist on Instagram or TikTok, it should feel like meeting them in person: knowledgeable, personable, expert, occasionally funny.
At Eater, the focus is on making internal talent more comfortable on camera, with editing and creating social-first content, according to Kaitlin Bray, senior director of audience development for Eater, Punch, Thrillist at Vox Media. “The medium of telling the story has shifted, but the desire to get their stories in front of as many readers still stands,” she says. “Even though the method has changed, I think there’s a lot of natural interest from staffers in sharpening skills or learning new skills.”
She says that Eater leverages advantages creators lack: built authority, unique access, and deep expertise. The effort has been adapting their voice to be more casual while delivering rigorous reporting.
Format adaptation and audience engagement strategies
Publishers tailor their approach based on each social platform’s audience and culture.
“We often find that our followers there are looking for an immediate and direct entry point into stories,” Benwell says. “On social, we’ve experimented with a lot of different formats. But what really resonates are those specific moments, where you kind of land in with a very specific person.”
He points to a recent success with original illustrations by one of his social producers for Naomi Beinart’s ‘new chill girl’ op-ed. “We did original illustrations for that, which is quite different from what we usually do. It was one of our biggest posts of this year.”
At The Verge, experimentation on social platforms is constant, according to Denise Cervantes, senior social media manager at The Verge. She describes how a recent conversation in a creative meeting about viral Sora videos led to a new video the same day. “We saw that and we’re like, okay, that is like a really good place for The Verge to show up,” Cervantes explains, adding, “The format of the video is just taking the viral AI video and having our host hop in and be like, ‘Hey, this is AI and this is how you can tell.’”
Eater encourages staff-driven experimentation with formats or ideas they’re passionate about. For example, a Gen Z social producer pitched an idea: she’d never tried steak tartare before and wanted to try four in one day, biking around Manhattan and Brooklyn, Bray explains. “That’s not a video concept that I would’ve come up with. It was rooted her in her experience and she ended up making this super fun video.”
Kate Scott-Dawkins notes that highly polished content still has its place, pointing out that massive audiences tune into premium shows like Wednesday. However, she notes that it feels out of place in spaces that started as social platforms. “It feels less genuine in these spaces that started out as social forms, off the cuff pieces of content,” she says.
Perhaps the most critical shift is treating every piece of content as potentially reaching someone who’s never heard of your brand. “Our philosophy really is to treat every Verge video thinking about an audience that doesn’t know who The Verge is,” Cervantes explains. “We can’t always go into it thinking this video is going to get served to all of our followers because of how the algorithm and discovery feeds work.”
Creating quality content for younger audiences and algorithms
While it’s a shift in strategy and fundamental mindset, the publishers we spoke to offered clear guidance for others working to reach younger audiences on social platforms. It requires leaning into personalities and staying on top of trends and algorithmic changes.
Benwell emphasized the importance of being data-informed rather than data-led.
“We have a saying here at The Guardian that we don’t follow the data blindly. We’re data informed, we’re not data led,” he says. “We make a real effort to be aware of the data and know what your key metric is. For us, it’s shares because that’s the metric where our work is being placed in front of more and more people.”
Authenticity remains critical, Cervantes stresses. “Audiences may not always want to connect with a brand, but there will always be that people-to people connection.”
Bray urged publishers not to be afraid of format experimentation while maintaining journalistic rigor. “We still have editors who used to write 1,000-word articles,” she says. “They’re still applying their journalistic skills, they are just figuring out different ways to communicate the information.”
The new landscape is clear: adapt to algorithmic distribution and creator-style content without abandoning your editorial standards. Newsrooms have found that experimentation and staff personality can coexist with rigorous reporting. Meeting younger audiences where they’re at doesn’t mean compromising good journalism.
For publishers, this means the path forward isn’t choosing between authenticity and polish, or between editorial integrity and platform fluency. It’s finding ways to deliver it all, in formats with voices that feel native to the spaces where Gen Z and Y spend their time.
Media product leaders are navigating uncharted territory in 2025, where generative AI is reshaping audience behavior and attention has become the scarcest commodity. Yet a consensus is emerging that the path forward isn’t about choosing between user needs and commercial success. Instead, it’s about aligning them.
Six senior product executives from Gannett, FOX, People Inc., The Economist, The Atlantic and The Wall Street Journal share how they’re navigating AI disruption, multi-generational audiences, competing for scarce attention, and aligning user needs with business goals.
Generative AI impacts everything
All of the executives DCN spoke with cited generative AI as the most disruptive technology shaping their product strategy. However, their thoughts and approaches varied.
“From a product perspective, we need to be sure that anything we do is beneficial to the reader,” explained Taneth Evans, head of digital for The Wall Street Journal. “One good example is our AI summaries at the top of articles. We first articulated a hypothesis, then we ran A/B tests and spoke to readers in order to make sure the WSJ audience was always at the heart of our ultimate decision.”
Adam McClean, chief product officer at People Inc. says generative AI is disrupting user behavior and audience acquisition, pushing media companies to diversify how they reach people. However, he points out that “It also opens up new opportunities to improve efficiency inside our organizations. People Inc. is no exception, and we are adjusting our strategy to meet both the challenges and the opportunities this technology creates.”
At The Economist, Chief Product Officer Liz Goulding says that generative AI is influencing both how people consume content and how they create products. “These two forces may look separate, but in reality, they reinforce each other,” she says. “As discovery habits shift, the need for our products to stand out… becomes even more critical. The tools available to product teams — from sharper analytics to generative design aids — are helping us respond to those needs faster and with greater precision.”
Kara Chiles, senior vice president of media products at Gannett | USA TODAY Network, explained the company is experimenting with automation and AI-based systems, anticipating it will have the greatest impact on their product strategy and business. “This isn’t new – we’ve been experimenting with this technology for some time. What is new is awareness across the board that this moment is different,” she says. “It’s opening up some bold thinking about how we create and sustain differently.”
FOX’s product teams are identifying where generative AI can enhance the consumer experience, help to personalize user journeys, and improve operational efficiency, explained John Fiedler, EVP of product & engineering at Fox. “We are being intentional. We’re prioritizing use cases that make our products more intuitive without adding unnecessary complexity. The goal is to integrate these technologies in ways that genuinely elevate the overall experience.”
Put users first
The media product leaders we spoke with consistently emphasized that putting audience needs at the center of product development didn’t compete with commercial goals, rather, it drove them. But how that alignment played out in strategy and practice varied across organizations.
Adam McClean and Liz Goulding emphasized strategic alignment, that user needs drive commercial value.
“If you are building products the right way, commercial objectives should enhance the user experience rather than compete with it,” says McClean. “At People Inc. we think of it less as a balancing act and more as an alignment exercise. When we design features that truly meet user needs, they naturally create value for our business.”
Whereas Goulding explains she’s always viewed user needs and commercial objectives as far more aligned than they’re sometimes portrayed. “Meeting user needs makes a product more desirable and reduces churn — that’s a win commercially. At The Economist, we approach this systematically. We assign tangible value to user needs within our internal prioritization framework.”
Other executives focused on practical execution, explaining how to implement that alignment through research, feedback, and iteration. Great product experience builds audience and brand loyalty, which drives monetization over time.
Gannett incorporates user needs at multiple points through new product creation, Chiles explains. “From identifying the task we are looking to accomplish, understanding the target audience, and validating whether the concepts we’re excited about are reaching the intended user. It’s challenging when you put something fresh and different in front of an existing audience – change is friction.”
Fox starts their product discovery process with the consumer in mind, then brings in commercial considerations to ensure every feature is evaluated for both impact and sustainability, Fiedler says. “Striking the right balance means building thoughtfully, listening to users, analyzing the data, and being ready to adjust if something isn’t delivering for both the audience and the business.”
At The Atlantic, the product team is focused on producing the best, highest quality reporting for their audience and delivering it in the most intuitive and creative way they can, says Gitesh Gohel, chief product officer at The Atlantic. “When we nail that product-market fit with The Atlantic’s values baked into everything we’re thinking about, we end up serving both sides really well,” he says. “We also know what we’re really good at – helping people understand the why behind the stories – and we don’t try to be everything to everyone. That focus is key.”
The biggest challenges in media product strategy
This year, media product leaders are facing complex challenges, from audiences spanning generations, to delivering seamless experiences, all without compromising quality or brand trust.
The Atlantic’s Gohel says they’re always exploring how they can expand what the publication can do for its readers or what it can help them with. “We’re spending a lot of time right now thinking about our mobile app experiences and building direct relationships with our readers. We’re thinking about moving our audience through all the different experiences and products The Atlantic has,” he says.
At The Wall Street Journal, Evans explains that their audiences and their expectations are shifting significantly. “We now span many generations with different preferences in how they want to consume our journalism, and we have to be there to meet them all. That means thinking about text and text-adjacent formats, as well as video, audio, graphics, conversations and more,” Evans says. “And the great challenge facing us all: how do we integrate all of this into a seamless experience fit for our many differing audience needs?”
Yet, for others, the challenge is about balancing speed, quality, and brand.
“We have to move fast to keep pace with the market, while still protecting the brand equity and product quality our users expect from us,” says McClean.
“For us, the biggest challenge is focus,” says Fiedler. “With so many emerging technologies, platforms, and potential partnerships, the most challenging hardest – and most important – decisions are often choosing what not to pursue. That discipline allows us to concentrate our efforts on what matters most to our audience.”
Another challenge product leaders spoke about was competing for audience attention and time, particularly as they consume news in so many places, says Goulding.
“Attention is scarce, and our audience has endless options at their fingertips,” she says. “For The Economist, the opportunity lies in deepening the connection between our journalism and our readers. [We must ensure] that when people choose to spend time with us, it feels not only rewarding but indispensable.” That means creating products that meet The Economist’s audience where they are, in formats that suit their lives, without compromising on quality.
It also means addressing the challenge of creating useful and meaningful content for audiences that empower audiences. “I can’t think of a time where having accurate and credible information has been more important,” says Chiles. “Our goal is to provide that information through an experience that makes users feel confident they have the tools to be informed, entertained and make better decisions for their lives.”
Design for people, teams, and purpose
For product executives in media in 2025, every decision ripples across audiences, teams, and platforms. Successful product leaders have discovered that true alignment, where great user experience drives revenue, isn’t just better strategy, it’s the only sustainable path forward. As audiences fragment across generations and platforms, and AI reshapes how people discover and consume content, the complexity facing product executives is undeniable.
However, these leaders have found clarity. When user needs drive everything, the challenge shifts from managing trade-offs to orchestrating alignment across strategy and business objectives. In a landscape where attention is currency, this isn’t just good product philosophy, it’s survival.
We’re not at Google Zero quite yet. But, as we near this point where Google search results provide direct answers and reduce outbound links, publishers face a critical imperative: They must build direct connections, maintain the loyalty of existing readers, and deeply engage audiences.
Since Google introduced AI Overviews 14 months ago, the AI-generated summaries have hurt publishers’ bottom lines, scuttled search traffic, and impacted ad revenue and subscriptions. This has resulted “zero-click” searches and a sharp decline in traffic, which some have dubbed “Google Zero”. News searches resulting in no click-throughs to news websites grew from 56% to nearly 69% as of May 2025, according to report from digital market intelligence company Similarweb.
About 40% of The Atlantic’s traffic comes from search, CEO Nicholas Thompson told Azeem Azhar, founder of Exponential View. “We’re seeing a significant decline, maybe a 20% decline,” he said. This translates to an 8% drop in overall website visitors, impacting ad revenue, subscriptions, and brand awareness.
AI Overviews appear in 39% of Google queries, according to Website Planet. An estimated 5.6% of U.S. search traffic on desktop browsers last month went to AI-powered large language models, according to market intelligence firm Datos, The Wall Street Journal reported. And analytics company Authoritas found that a site previously ranked first in a search result lost 79% of its traffic if results were delivered below an AI overview.
Whether publishers like it or not, that traffic isn’t coming back. Google Zero looms large on the horizon. Pew Research found that for searches with AI summaries, Google users clicked on a traditional search result link in only 8% of visits, compared to 15% without AI summaries.
And plenty of publishers aren’t happy about things. A legal complaint was submitted to the UK’s Competition and Markets Authority over the impact of Google AI Overviews on news publishers, arguing that Google is abusing its market dominance by using publisher content in AI-generated responses without fair compensation, while simultaneously reducing traffic to news websites. Additionally, The Independant Publishers Alliance filed an antitrust complaint with the European Commission in June, alleging that Google abuses its market power in online search.
Meanwhile, as publishers adapt to this new reality, DCN spoke to some industry experts to reveal what’s working.
Direct relationships important, valuable
Some publishers aren’t waiting for Google Zero, they’re already building the direct audience relationships and cultivating reader loyalty that will matter even more when search traffic disappears.
Hearst Connecticut Media Group strategically builds direct audience relationships as part of its long-term audience strategy. Their GameTime CT high school sports vertical, which serves high school parents, athletes and coaches, drives subscriptions according to Mandy Hofmockel, managing editor of audience at Hearst Media Connecticut.
“It goes above and beyond game coverage and is a formula for success in serving local audiences,” Hofmockel says. “Our high school sports coverage is evolving to be less about the sport itself and more about athletes, their stories around the game. We are building depth, experiences and connections.”
To deepen engagement, the company launched it’s first texting campaign for UConn basketball during March Madness this year, she says. “That platform allowed that small but engaged pool of UConn fans to message us directly with questions on everything from player injuries to coach strategies, to just like where they could watch the game,” she says.
Hearst Media Connecticut is also deepening its reporting in key local areas, including weather, education and real estate, even adding a meteorologist to its team and developing weather tools and trackers. Hearst’s hyperlocal approach, covering school closures, local weather patterns, and community-specific issues, provides indispensable information creating direct traffic that survives the death of search.
Being a local publisher provides advantage. “We know what it’s like to live in, go to school in and eat across the state. That’s reflected in our coverage and the key coverage areas for the newsroom,” Hofmockel says. “Our readers don’t hesitate to share us with us what they think of our coverage because they feel that connection to us.” This means when readers need to know what’s really happening in their town or city, they come straight to Hearst instead of searching for answers.
Direct engagement, across multiple platforms
BBC Studios has “embraced all manner of platforms to reach audiences wherever they are, from a thriving BBC News WhatsApp channel to Instagram clubs for our Culture super-fans.” However, their primary focus on building stronger, direct relationships through product innovation and editorial strategy, says Ben Goldberger, GM and executive director of editorial content. “Central to this is the relaunch of our BBC.com site and app, which offers a premium, more streamlined user experience that encouraged repeat visits and deeper engagement,” Goldberger says.
The company rolled out a new pay model on BBC.com in the US, which Goldberger called an important step in strengthening the connection with their most passionate users. The launch of the new pay model will help them gain greater insight into their audiences’ preferences and behaviors, and to strengthen those connections.
“One constant has been our commitment to our owned-and-operated channels,” he says. “We have seen meaningful success driving engagement on our platforms as we reduce reliance on those of others.”
BBC Studios operates 11 regular newsletters for global audiences, spanning topics from US politics to personal health. Goldberger says the response to these has been incredibly encouraging. “Our newsletters subscribers are deeply engaged with our work, regularly visiting BBC.com from newsletter links and taking the time to send thoughtful, considered feedback,” he says. “Indeed, the outpouring of notes from readers of our In History newsletter led us to create a recurring section featuring reader memories that is among the most popular.”
BBC Studios aren’t just creating an email list, they’re creating a feedback loop where engaged newsletter readers become content contributors and reliable traffic drivers. It’s a two-way relationship that generates both audience loyalty and editorial material. BBC Studios owns every touchpoint in the reader journey, making them insulated from external platform disruptions like Google Zero.
Building stories AI can’t replicate
David Skok built The Logic, a Canadian business publication focused on technology and innovation, as a subscription-first publication from launch in 2019. “Our business model necessitated us having a direct relationship with our readers from the start,” the CEO and editor-in-chief says.
In a recent column about AI’s impact on journalism, Skok delved into the existential question facing publishers as AI upends traditional web discovery. He believes in creating stories that no AI platform can summarize accurately.
“If you’re writing stuff that is yours, exclusively yours, you cannot get anywhere else and isn’t answered in just one pithy response from a chat engine, that’s how you’re going to win,” he continues. “The thing that’s really still going to differentiate you is what stories are you assigning and what stories are your reporters pitching? Are they things that you cannot get anywhere else?”
Beyond content strategy, The Logic is intentional about their audience engagement. “I think intimate events are really working,” Skok explains, describing some of The Logic’s recent events. “We’ll go to a place like Calgary or Vancouver and have 30 people for breakfast and just talk about the issues of the day, bring in one of our columnists, and those kinds of things are extreme value for a smaller group of people. And they feel it.”
The Logic also has a Slack channel for direct engagement with reporters, and hosts virtual events based on breaking news. “We try to make sure that our readers understand that what they’re getting with The Logic subscription is way more than just access to ungated content behind a paywall,” he says.
Beyond the (zero) click
“Publishers will have to shift their expectation expectations from some of the primary referral sources we’ve relied on in the past,” Hofmockel says. “It doesn’t mean we completely give up on search and social. But we have to adjust our strategies and find additional ways to connect with our communities.”
As publishers focus on direct relationships, they continue to make sure they’re “maximized for visibility,” on Google because “it’s still an important channel for distributing our work,” Hofmockel from Hearst says. However, it is important to take a strategic, rather than dependant, approach. In other words: never be solely dependent on platforms for your core business model.
Hofmockel believes it’s an opportunity to reevaluate not just audience strategies, but publishers’ content approach. “We can build new, distinctive products (that) are rooted in data, go deep in the categories that matter… and make sure we’re giving readers reasons to come and subscribe. Building around these needs with expert reporting will make us essential with or without platforms,” she says.
Building direct relationships, on whatever platforms you own, whether it’s newsletters, events podcasts, or content verticals, publishers must be conscious and intentional about owning their audiences, according to The Logic’s Skok.
“I think that’s the most important relationship you can have, and it’s the one that will allow you to withstand this change. The thing with a subscription business, like a paywall business like we have is, it’s so much harder to build it up. It’s slower, it’s more methodical. There’s no quick hack, growth hack to make it happen. But once you’ve built it up, it’s really hard to tear it down because these readers are invested in your success.”
As AI reshapes digital discovery, publishers who cultivate direct, meaningful relationships with engaged audiences, position themselves to survive and thrive in the post-Google Zero landscape.
The 2025 DCN Next Summit kicked off in Miami April 22 with an energizing atmosphere as senior media executives from DCN’s member companies came together to discuss the biggest issues and opportunities impacting the future of media.
In his welcome, DCN CEO Jason Kint highlighted the challenging environment the media finds itself in. “Let’s be honest, the last 12 months have been volatile,” Kint said, “And the volatility isn’t just economic, it’s institutional. The forces testing our economy are also now testing our democratic norms, including a free and plural press itself. [We face] a direct challenge to the independence of the press and the principle that journalists, not governments, get to determine the language of truth.”
This, Kint said, is the new normal: accelerated pressure, relentless power grabs and heightened scrutiny all at once. “It’s messy, it’s uncomfortable, and it’s redefining the rules that we all play by.”
In the midst of this, Kint highlighted premium content still matters but what defines it is changing. “Growth is harder, but it is possible, especially as you strengthen your direct relationships with your audience and customers. Trust… is everything. It’s foundational and it must be defended. And, in times of vulnerability is when you build on it.”
While the topics of discussion both on stage and off were wide-ranging, three significant themes emerged: the importance and evolution of trust, the value of direct audience relationships, and new influencer dynamics impacting media brands.
Trust in a fragmented world
In an era where audience attention is fragmented across numerous platforms, trust is the core value exchange between a media brand and its audience. Katherine Maher, president and CEO of NPR, emphasized the importance of maintaining editorial independence and impartiality as essential components of trust.
Katherine Maher, president and CEO of NPR
She said, “Our editorial independence is paramount. People listen to NPR and they care about public media because they trust it and they know that it is independent. To my mind, if we cannot maintain that editorial integrity, we cannot serve our audiences the way we need to be served.”
This foundational trust faces new challenges. New research from DCN and Magid on Gen Z’s video consumption reveals a significant difference in trust levels between individual creators and brands, with individual creators generally being perceived as more trustworthy. The study, called “Decoding Video Content Engagement,” talked to 1,000 young people aged 13-40, to understand how they saw media brands. The results (available to DCN members) suggests that Gen Z’s understanding of what is trustworthy is evolving based on where they spend their time and energy.
“When you talk to Gen Z, it’s the individual that’s most valued. It’s the influencers, it’s the streamers,” Andrew Hare, SVP, head of quantitative research at Frank N. Magid Associates explained to attendees. Media companies face a significant challenge in building trust with Gen Z and Gen Y, and being seen as trustworthy, authentic and interesting, compared to individual creators, who are overwhelmingly trusted more by these generations.
Hare mentioned an opportunity for digital media companies to “collaborate and co-create with creators themselves to maybe even add some trust back to the brands.” He noted that digital media companies must focus on humanizing their brands, fostering direct relationships with audiences, and finding ways to be real and relatable while upholding their journalistic standards.
The evolving role of creators
Discussions at the summit frequently touched upon the evolving role of journalists in today’s media landscape and the rise of individual creators/influencers as a force in news. According to a November 2024 study by the Pew Research Center, 21% of U.S. adults now regularly get news from influencers. This figure rises to 37% among those under 30—an age group that is increasingly difficult for traditional outlets to reach.
Tiffany Sam Chow, SVP, strategy and business development at NBCU News
Tiffany Sam Chow, SVP, strategy and business development at NBCU News Group, pointed out that news anchors are becoming personalities on platforms like TikTok, which allows them to build individual connections with audiences. This shift changes the role of anchors from authoritative figures to relatable personalities, she explained.
Chow cites the example of Savannah Sellers on TikTok. “She does these behind the scenes where people can understand her as a person,” Chow explained. “People start following her on social as a person and then start following her on social as a news anchor.” As people engage with the on-air talent on a personal level, they begin following them as journalists, and in turn, engage with the NBC News and Today Show handles, Chow said.
Sam Felix, SVP, Strategic Partnerships & Business Development, at CNN echoed this shift. She noted CNN has also been thinking about how to drive that relationship between their on-air talent and audiences. “Part of our superpower is our ability to produce video at scale and this amazing talent. We have the right ingredients to engage with this audience. But we have to figure out (how) to pull back the curtain, get them sort of like closer, one-on-one, with this audience in a way that they seek us.”
In addition to their shows, CNN personalities produce multiple vertical videos per day, published on social channels and on CNN’s platform, Felix said. “Over the next several months, as you see the kind of next phase of CNN come out into the world, you’ll see that same type of production format be at the center of the content and our products, because it is resonating.”
MLB’s VP, Social Media and Innovation Cameron Gidari noted that some baseball creators are as popular, if not more so, than baseball players “kids are recognizing them!” Thus, their strategy involves empowering these creators. “We have a really robust crop of up and coming baseball creators,” Cameron. “They’re non-traditional media for a new age.”
MLB’s creator strategy involves helping empower creators, to help them grow, giving them access to events and sharing their content. “We went to help them grow because we know that they’re Baseball Tonight for the next generation, right?”
Building deeper connections with direct relationships
Publishers have long held direct relationships with audiences, built on trust and high-quality content. These relationships allow media companies to understand and anticipate audience needs. Strategic insights also inform monetization strategies like subscriptions, events and advertising.
In 2025, strengthening direct relationships with audiences has never been more critical. As media companies expand beyond traditional advertising into licensing and other D2C strategies, deepening audience connections is essential for sustainable growth.
Daniel Alegre, CEO, TelevisaUnivision
CEO Daniel Alegre credits his company’s success to TelevisaUnivision’s vast Spanish-language content catalog, built over 80 years, which helps nurture a direct, multi-platform relationship with audiences. TelevisaUnivision integrated its operations and created a single content strategy that serves linear TV in both the U.S. and Mexico and ViX, its streaming platform.
Alegre noted that the company continues to innovate in video content to engage new audiences. They are developing one-minute “micro telenovelas” specifically designed for mobile consumption. “These are essentially made for the phone, and can create new commercialization opportunities for subscription and advertising … We can also work on microtransactions,” he said.
At the Athletic, Publisher David Perpich explained that the company is exploring partnerships to leverage its content and audience, including a partnership with MGM which integrated betting coverage, and Stubhub which allowed users to purchase tickets within The Athletic’s content.
And in a move that is certain to be a fan favorite, MLB formed a “partnership with eBay where we have a collectibles vertical and you can buy on eBay,” he said. The focus of these initiatives is on “how do we create content that consumers would love but then let’s figure out the right business model on the other side to take advantage of it.”
Relationships are also changing between media companies, brands and advertisers, with a greater emphasis on direct relationships and mission alignment.
Shannon Watkins, CMO, Fiserv
Shannon Watkins, chief marketing officer at Fiserv, explained that Fiserv increasingly bypasses media agencies, instead partnering directly with media companies, viewing them as extensions of their own marketing team. This direct model allows Fiserv to keep strategy development in-house while collaborating with media partners to execute.
“It’s less about the dollars and cents and more about that symbiosis that you can have with your partner media or otherwise, where it is a true mission alignment because then the conversation moves beyond placements and dollars, but how can we grow together? And that’s what we’re looking for,” she said.
Persevering and pushing forward
As digital media companies grapple with the challenge of maintaining trust amid increased scrutiny and competition from more personalized, often more relatable creators, the importance of direct, authentic relationships with audiences has never been clearer. Media are learning to adapt to this shifting landscape, where collaboration with creators can help rebuild trust while still maintaining journalistic integrity.
Media companies must evolve to stay relevant. However, they must also safeguard the foundational values that have long underpinned their role in society, including press freedom. This Summit highlighted how they are persisting through challenges. As Kint pointed out, “We must keep pushing for fair value, for IP protection, for a level playing field, in equal competition. And above all we must defend the role of a free and plural press at a moment when institutions are being tested from every angle, even at the highest office in the land.”
Content licensing has long been an important revenue stream for digital media companies. For decades, it allowed publishers to monetize their content by granting rights for others to republish or repurpose their material, evolving from licensing to aggregators, databases, social platforms, to streaming video services. Now, content licensing faces another evolution: artificial intelligence (AI).
Digital media publishers are finding themselves in a unique position in that they possess decades worth of quality content AI companies crave. “Over the next few years, content creators and AI companies will deepen their relationships,” predicts Yulia Petrossian Boyle, founder and principal of YPB Global LLC and FIPP chair. “However, as AI players try to secure more original content, those relationships will need to transition from one-off deals to well-structured, ethical partnerships with strict IP protection and meaningful ongoing revenue for publishers.”
TIME’s COO Mark Howard believes that publishers currently have three ways they can approach the AI dilemma: “You can do nothing. That’s just not something we would consider, to sit on the sidelines and just let everybody else figure it out. The other two options are to litigate and negotiate. Litigation is a very, very large commitment… So, that leaves negotiation.”
For some media companies, AI licensing agreements offer an alluring mix of copyright protection and monetization opportunities as DCN contributor Damian Radcliffe points out. And, as they negotiate these deals, publishers are discovering they must balance the potential for monetization with the need to protect intellectual property rights, navigate complex legal challenges, and ensure responsible AI usage.
Fair value in AI content licensing
According to a recent INMA report, executives considering licensing deals need to understand the value of their content in an AI-driven market. Then they have to negotiate attribution and compensation models that align with business goals. The report recommends collaborating with industry peers to create standardized agreements. It emphasizes the importance of advocating for responsible AI practices, including transparency in data usage.
Image credit: Ezra Eeman, Strategy & Innovation Director – NPO
The report also highlights emerging licensing models, which include direct licensing, value-in-kind partnerships, training fees, bundled partnerships, and per-use compensation. Boyle notes promising approaches, like “data-as-currency” deals, where AI companies offer analytics in exchange for access to their platforms and services (in some cases in addition to some smaller flat fees).
“Revenue-sharing is on the rise, where publishers earn a portion of subscription revenue or performance-based compensation (based on lead-gen, or engagement analytics),” she says. “For example, Perplexity AI’s Publishing Program launched in July 2024 offers revenue share based on the number of a publisher’s web pages cited in AI-generated responses to user queries. Those in the program earn a variable percentage of ad revenue generated per cited page.”
Boyle says that, while compensation models are improving, she worries that AI companies do not adequately compensate for content that has higher production costs, such as investigative journalism. She points to pushback from publishers like Forbes, who rejected the Perplexity proposal.
Negotiating with AI companies on behalf of her consultancy, Boyle has observed that offers by some AI companies for training datasets are insufficient. “Since agreements are not indefinite, it is unclear to me how publishers will be compensated in future when AI companies may no longer need training data for their data sets.”
In her opinion, current compensation models between major AI companies and publishers do not adequately reflect the significant investments that publishers make in creating original content. She believes compared to the substantial amounts AI companies invest in technology, such as chips, their expenditure on content seems disproportionately low. This disparity highlights a need for a more balanced financial recognition of the value that original content creators bring to these partnerships, she says.
However, striking these deals isn’t simple. Howard notes that each one is different, each has different monetization models and philosophies on revenue sharing.
“Some of them are flat fee for training, some of them are variable based on user adoption of their own products, and some of them are based on future ad models that haven’t even launched yet,” Howard says. “Many of them have some form of value-in-kind around technology or technology resources, which makes me very excited. I think that that may end up being where most of the value is derived in the long term.”
A few of TIME’s AI partnerships are infrastructure-based, like Fox Verify, which uses their blockchain-based technology to verify all of the content TIME publishes in the CMS. This provides them with a ledger of all of their intellectual property going forward. After that, according to Howard, they worked with Tollbit and Scalepost to track and monitor all of the AI bots on TIME’s site any given day and see what they’re doing.
Access to technology is a key benefit of TIME’s AI partnerships for Howard. “We’re partners of theirs. I have direct access to their CTO and their senior leadership team. We get to hear what… they’re thinking about the market, that’s a really valuable conversation for us to have.”
“We brought money in as a result of these deals,” he says. “I’m happy about what we brought in. Some of it is fixed, a lot of it is variable and a lot of it is access to product resources and technology.”
Factiva puts trust first in its AI licensing
Dow Jones launched Factiva Smart Summary in November, a groundbreaking feature in its business intelligence platform engineered with Google’s Gemini models on Google Cloud. Smart Summary leverages generative AI technology to create concise summaries for Factiva users that are fully transparent and traceable, utilizing licensed content from each of their publishing partners.
To do so, Factiva approached every one of its nearly 4,000 sources in 160 countries with licensing agreements. “We did this because we are a publisher first and arbiter for publishers… We won’t ask any of our publishing partners to do anything that we’re not prepared to do ourselves,” explains Traci Mabrey, general manager of Factiva. “As such, we have elected and will continue to elect, to reach out to publishing entities and request additional licensing permissions and actual rights for generative AI use.” Today, its marketplace includes nearly 5,000 partners.
Dow Jones emphasizes the importance of respecting and compensating intellectual property and content creation. Mabrey outlines four key criteria guiding their AI partnerships: trust, transparency, segmentation, and compliance.
“We believe that trust is imperative. We believe there needs to be transparency in terms of content being created, used, surfaced and attributed,” Mabrey says. “There also needs to be relative segmentation in terms of use cases across different solutions. And there needs to be compliance and governance to adherence to the first three, of trust, transparency and segmentation.”
Deal points when licensing content for AI training
There’s no one-size-fits-all model for licensing deals, and the best approach depends on a publisher’s specific goals, content, and resources. Some determine how easily an LLM can integrate into their existing systems and CMS. Some choose LLMs based on those they already deal with.
But, data privacy and security are central concerns in these agreements. Vadim Supitskiy, chief digital and information officer at Forbes, told Digiday that ensuring interactions with AI products remain safe and protected is a key priority.
Mabrey echoes this sentiment, emphasizing that privacy and security are integral components to negotiations with AI partners. “As we’re looking at responsible delivery of AI, responsible usage of content and privacy and security in terms of technical infrastructure, that is our leading indicator.”
Publishers must have review rights over AI-generated outputs, ability to see proof of usage logs, and be able to enforce brand guidelines, according to Boyle. “All those things have to be clearly defined in the licensing agreements. Tracking metrics of engagement, attribution, and demographic insights is also important for publishers to receive, to be able to see how valuable their licensed content is,” she says.
Essential safeguards in the agreements themselves ought to include strong, sophisticated clauses to protect publishers’ IP, says Boyle, “including mechanisms to prevent unauthorized reproduction, clear ownership definitions, restrictions on data usage, well defined termination provisions, attribution and fair compensation.”
Howard emphasizes that no two content licensing deals with AI companies are the same, and each comes with significant legal and technical hurdles. “First, there’s the legal aspect and every company needs to come up with their own legal terms and what is acceptable to them and what is not. What do they have the rights to? What do they not have the rights to?” he says.
“Once you’ve determined all of that, you need a technology solution to be able to deliver the content to them… All of the delivery mechanisms are quite different and require some form of customization.”
These complexities point to why AI companies have slowed the pace of new licensing agreements after an initial rush. Negotiating unique terms and building tailored tech solutions for each partner has proven difficult to scale, Howard notes.
Where AI licensing is headed
AI is reshaping how content is distributed, discovered, and monetized. For media companies, the choice is clear: engage in legal battles or proactively negotiate terms that ensure fair compensation. The market is rapidly evolving with new players, technologies and partnership models.
For companies currently negotiating content licensing deals with AI, Howard says to move forward. He points out that, while there are benchmarks based on what other companies have secured, the initial rush of deals has likely passed. He doesn’t expect future deals to improve; in fact, he thinks they’ll probably get worse.
Mabrey believes that the industry has reached a unique inflection point, where generative AI gives it the chance to assert that content is intellectual property and requires compensation. “We, as a media community around the world, should be coming together to assure that all of us are asserting our rights in the same manner.”
In light of these shifts, there’s a clear message for media executives: the future of content licensing is in their hands. Instead of letting the industry define them, publishers can shape the future of the industry by hammering out a windfall through litigation and the courts, negotiating partnerships, and advocating for fair treatment.
As news consumption evolves, The New York Times’ R&D Lab department is at the forefront of redefining how stories are told. Their goal? To enhance the narrative experience without sacrificing journalistic integrity.
With a focus on exploring how emerging technologies can be applied in service of journalism, the department is proving that innovation can enrich journalism rather than distract from it. But it’s a fine balance. In a landscape littered with shiny new things, where innovation is often considered a buzzword, the R&D Lab team pioneer new technologies to serve the story.
Scott Lowenstein, director of research and development strategy at The New York Times, oversees the R&D team, leading experimentation with the next generation of storytelling. The R&D group is a mostly technical team that works on applications of emerging technology in the service of journalism, but also for their products and business, he explained.
From the NYT’s R&D Lab’s experiments in mixed reality journalism.
The launch of virtual reality (VR) in 2016 marked a significant achievement, as the R&D Lab created over 30 immersive films that showcased the narrative potential of VR. The films received high viewer engagement marked by over 1 million downloads of their VR app. In 2017, the team supported newsroom storytelling by integrating emerging technologies like mixed reality and connected home devices, while also improving public figure identification in photos through automation.
In 2021, the team developed the Times’ first AR game and animated sequences for augmented reality, continuing to explore spatial journalism and improve media transmission with 5G technology. By 2022, they emphasized integrating locative data and spatial understanding into journalism, aiming to better reflect and connect with communities, thus enhancing storytelling.
“Never ending cycle of exciting, interesting problems to tackle”
A lot of the R&D team’s work starts when Times’ staffers come to them with an ambitious project they’d like to do, but don’t know where to start. Lowenstein said the team does their research on technologies, the latest tools in the market, and projects in adjacent industries from architecture to game design.
“We prioritize the ones that feel achievable. We find the technology to match the ambition of the story,” Lowenstein said. “We will oftentimes say, this technology is really cool, but it’s not ready. But, when we do find an idea that works and match it well with the story, then we’ll help that team come up with the first execution.”
And, when the team finds a technology that works, they’ll find ways to make it repeatable, so that future stories within The Times can use it. “We’ll also often open source our work and share it with a broader journalism community to help replicate what we’ve done.”
spatial journalism in a broad sense, which encompasses things from 3D to mixed reality and virtual reality.
“The mission is to try to use these tools in service of great storytelling and not try to do something gimmicky. Instead, we want do something where this new tool could help us do what we couldn’t do before,” Lowenstein said.
A lot of the projects are seriously ambitious. They’re not only hard to do, but they also require a large team of technologists, editors, and resources. So, the Lab is invested in how to facilitate this process.
“A lot of our ambition is trying to just make that whole process easier for the teams so that they can just say, we want to do this and that they can do it the same day,” Lowenstein said. “Some of it is building software and templates and things that don’t require a super-specialized 3D engineer to build a 3D story. That’s where a lot of our effort goes: finding those repeatable things and creating the tools that allow editors to do that work with less or no help from us.”
Some of Lowenstein’s favorite stories were during the early days of augmented reality. “We built a lot of stories that we published as 2D stories on the web, and then translated specific slices of them into AR, that people could use on their phones. Some of those were very well received and we learned a ton from that experience.”
Balancing innovation and integrity
The New York Times R&D Lab is not interested in employing the latest technology just for the sake of doing it. Technology is a tool in the service of their mission, not an end in itself. “One of the great things about working with our amazing journalists and editors is that their gimmick meter is really strong,” Lowenstein said.
Timeline of innovation at The New York Times’ R&D Lab
2006 – Launch of R&D Lab
Foundation: The New York Times R&D Lab was established to innovate and experiment with new technologies and storytelling methods.
2009 – Multimedia Applications
Commuter App: An innovative app that combines traffic cameras, Google Maps, and location-specific Times content, enhancing news delivery through geocoded articles and blog posts.
2011 – Interactive Technology
Magic Mirror: A data-bearing mirror using motion sensing and voice recognition to deliver on-demand information via the Times’ APIs.
2014 – Semantic Listening
New Tools: Developed systems for extracting semantic information and crowdsourcing data on cultural artifacts.
2015 – Wearable Technology
Focused on wearables, examining their potential uses in the media landscape.
2016 – Virtual Reality (VR)
Created over 30 immersive films.
2017 – Emerging Technologies
Shifted focus to support newsroom storytelling by integrating technologies such as mixed reality and connected home devices, while also improving public figure identification in photos through automation.
2019 – Photogrammetry, Spatial
Computer Vision: Revealing hidden narratives within visual content.
Photogrammetry: Creating immersive 3D environments from 2D photos.
Spatial Computing: Exploring new storytelling formats in augmented reality.
Media Transmission & NLP: Enhancing photo and video transmission and extracting insights from the NYT’s extensive archive.
Developed the Times’ first AR game and animated sequences for augmented reality, while continuing to explore spatial journalism and improving media transmission with 5G technology.
2022 – Spatial Journalism
Emphasized integrating locative data and spatial understanding into journalism to better reflect and connect with communities, thereby enhancing storytelling.
Technology evolves rapidly, and having a solid understanding of emerging tech applications is incredibly valuable. The R&D team began exploring large language models back in 2019. Now that these models are widely adopted, they have a significant advantage in knowing how to apply them responsibly.
From the NYT R&D Labs’ work on media transmission and provenance.
Lowenstein emphasized the importance of transparency in their work. “We’re always going to tell our readers when and how we’re using these emerging technologies,” he said. “I view it as a duty to our readers to explain how we’re using these technologies. And in that process of explanation, that is the process of balancing storytelling and integrity in action.”
Those principles show up in reader-facing ways. This is evident on their website, where stories include methodologies, labels, and sometimes tutorials or explanatory text. Those principles also show up in industry-facing ways. The R&D Lab shares their work as much as they can on their website, and through open source libraries for others in media.
“I think that’s one of the great benefits of having this kind of open source mentality is that we can see, we put our stuff out there and we can see how the community is using it,” Lowenstein said.
A lot of people are using the photogrammetry, the spatial journalism and 3D storytelling, using the libraries and contributing to them, he explained. “It’s super gratifying to see this little community built up around some of the photogrammetry, the spatial journalism. It’s awesome to see how people have taken that work and evolved it and made it their own.”
From the NYT’s R&D Lab’s experiments in spatial journalism.
Technology should be in service of the story
As technology and journalism continue to evolve, other organizations can glean valuable insights from The New York Times R&D Lab’s approach to experimenting with emerging technologies. For one, Lowenstein emphasizes that technology should be in service of the story. “That’s first and foremost. We don’t do it for the sake of doing it. The story should come first and you should really look through if the ambition of the story warrants it.”
Embracing experimentation is equally vital, and even though trial may lead to error, each misstep could lead to unexpected insights. “We often try things that don’t work the first time or the fifth time, and then you find the perfect fit,” Lowenstein said.
Lowenstein suggested keeping a record of what you’ve tried and what you’ve learned along the way. He says that oftentimes, things will pop up you didn’t know would be useful in the future. Keeping records allows you to adapt to new possibilities.
Finally, the director suggested that it is important for media companies to have principles about how they use technology responsibly and in service of their work. It helps guide decision making.
By establishing clear principles for the responsible application of emerging tech, organizations can carve out a focused space to innovate while delivering quality journalism that resonates with audiences. Ultimately, it’s this blend of technology, experimentation, and integrity that will empower the industry to navigate into the future.
“In some ways, having those clear principles about when and how you use technology is a great creative constraint. It allows you to define the parameters of how you want to use the technology in service of quality journalism. And, working within those boundaries from the beginning makes it achievable and makes the space of things that you could do smaller so that you can actually focus on the right stuff,” Lowenstein said.
It’s been a rough year for publishers and broadcasters in Canada. In the wake of Meta’s news ban, Canadian news publishers and broadcasters have faced declines in online traffic, engagement, and revenue. As the publishers we spoke to illustrate, the ban, implemented last August in response to Canada’s Online News Act, has dramatically altered the media landscape – and not for the better.
A recent study by the Media Ecosystem Observatory underscores the profound effects of these changes on Canadian media. The report, which examined the effect of Meta’s ban on Canadian news, revealed that news outlets lost 85% of their engagement on Facebook and Instagram, leading to a total engagement decline of 43%.
“Canadian news organizations lost an enormous amount of their online viewership,” explained Aengus Bridgman, director of the Media Ecosystem Observatory and co-author of the study. “They’ve been heavily relying on the distribution channels of Google and part Meta to get their work out there.” The study found that Canadians consumed less news as a result of the ban, and that Canadian news outlets saw a reduction of 11 million views a day.
The report, titled Old News, New Reality: A Year of Meta’s News Ban in Canada, found that approximately 30% of local news outlets in Canada are now inactive on social platforms. “Their digital presence has basically collapsed and they’re no longer connecting to the social web,” he said.
Navigating a power struggle
Meta’s news ban in Canada underscores the current struggle and challenges news media face amid the power of the platforms. Facebook began blocking Canadian news on its platform and on Instagram, in response to Canada’s Online News Act, which passed in June 2023. The legislation requires big tech companies like Meta and Google to pay media outlets for the news content they share on their platforms.
Rather than pay, Meta simply blocked Canadian news sharing on its platforms altogether. Within weeks, News Media Canada, the Canadian Association of Broadcasters and public broadcaster CBC/Radio-Canada filed an antitrust complaint against Meta accusing it of abusing its position.
However Meta claims that complying with the legislation would require them to pay for links to news content. Thus, they said they didn’t want to operate in what they considered an unfair regulatory environment.
“(The) legislation misrepresents the value news outlets receive when choosing to use our platforms. The legislation is based on the incorrect premise that Meta benefits unfairly from news content shared on our platforms, when the reverse is true,” the company said in a statement. “News outlets voluntarily share content on Facebook and Instagram to expand their audiences and help their bottom line. In contrast, we know the people using our platforms don’t come to us for news.”
Meta argued that their platform provided significant value to Canadian news publishers by driving over 1.9 billion clicks to their content in the year leading up to April 2022. The company valued that at $230 million in “free marketing.”
Meta’s decision in Canada recalls a similar situation in Australia in 2021, where it briefly banned news on its Australian platforms in response to the threat of regulation under that country’s News Media Bargaining Code. More recently, Meta stated that it will not renew any existing deals with Australian news outlets, nor those in France and Germany. In other words, it looks like Meta is taking a global stand against supporting news on its platform.
Impact of platform news bans on publishers
In Canada, the impact of the news ban was immediate. Traffic from Facebook plummeted, leading to a sharp decline in audience reach and digital ad revenue. Many Canadian publishers, especially smaller and independent outlets, faced increased financial strain as they lost traffic.
Chuck Lapointe, CEO of Narcity Media Group, saw years of investment disappear because of the ban. Since 2013, his online outlets have thrived as “social-first” platforms, relying heavily on Facebook and Instagram for distribution. Meta’s abrupt ban severed Narcity and MTL Blog’s connection to Canadian audiences, cutting off revenue streams tied to branded content deals and partnerships.”Right away, all of them were blocked and Canadians were not able to see our content. That was an immediate hit to revenue,” Lapointe said.
Facebook pages for Narcity and MTL Blog, which had well over 3 million followers, went dark. Lapointe watched as his audience dropped as a result. “We lost 50% of our traffic overnight, which impacted perception and trust.” He also found that “some writers and journalists said that they didn’t want to work with us because they can’t share their work openly.”
Great West Media, which produces 20 print publications across Alberta, also saw traffic and revenue decline. “If we look at traffic as the prime factor in driving revenue, then yes, the Meta news ban had a serious impact,” said Brian Bachynski, president of Great West Media.
Despite the Meta news ban, Great West has still managed to grow its audience online. The company is using a variety of methods, including a strategy for breaking news, newsletters, SEO and more. However, Bachynski said “I’m certain if Meta hadn’t banned news in Canada, our numbers would be greater.”
He has concluded that it is in no business’ best interest to rely heavily on a third party as the primary driver of revenue. “Because ad volume is predicated upon page views, the Meta ban has depressed traffic to our websites and therefore our ability to drive revenues. These revenues, however, are not enough to operate a plausible business,” he said. “Even without Meta’s ban, our audience isn’t large enough to exist on the extremely low ad rates carved out by the monopolistic practices of Meta, Amazon and Google.”
The decline in traffic appears to have been part of a longer-term trend in Canada, according to David Beers, editor in chief of The Tyee, a B.C.-based independent, online news magazine. Even before the official block, he watched The Tyee’s Facebook engagement dropping over the previous year—evidence that Facebook was already adjusting its algorithms to deprioritize news.
“We saw it. It was unmistakable. We compared notes with other publications. They saw the same,” Beers said. “It really throws a lot of doubt on Facebook’s claim that if it weren’t for the Online News Act, they would be supportive and sharing a lot of news on their site.”
Beers said that Facebook may be using Canada as an example to deter similar regulations in larger jurisdictions like California and the European Union. However, he said, “no one pays attention to Canada, so there’s no price to be paid by depriving us of the sharing of news.”
Meta decides what’s news
The ban rollout was also messy because it affected media outlets that don’t produce news. Among them was the satire and parody publication, The Beaverton, and campus radio stations, from CIUT 89.5 at the University of Toronto to CJSW 90.9 at the University of Calgary.
CJSW Station Manager Adam Kamis believes campus radio stations were included in the list of news agencies simply because they are community media, rather than being assessed individually. “We lost our Instagram and Facebook about this time last year,” he explained.
CJSW, which has been broadcasting since 1957, has 110 music, spoken word and multicultural programs in more than 10 languages. They have volunteer hosts talk about stories that matter to them, Kamis explained. “And, we don’t even really do news reporting that much.”
“Now we’re getting no money and we have no use and access to these services. And though our presence is rooted in the auditory experience, we live in a world where the visual needs to be part of the presence as well. Without that, it’s like we have the worst of all possible worlds befalling us where we have no presence and no money.”
As a campus community radio station, CJSW relies on community and listener engagement. They leverage their following for ads and sponsorship. The ban impacted the station’s ability to sponsor local music festivals and its annual funding drive, which generates essential revenue to support station operations, programming, and services.
It’s usually an amazing opportunity to share on social media, but not with a ban in place, Kamis said. “Last year was our first year without it,” Kamis said. “We were able to leverage our connections with our friends of the community to help… but that’s a thing I don’t want to rely on year after year.”
Meta’s interpretation of what qualifies as news posed a particularly ironic challenge for digital outlet Narcity–one that demonstrates the subjectivity of the company’s decisions around what to block. “The government doesn’t hold an official list of news organizations that Meta would have to comply with,” said Lapointe. “So Meta… relies on their interpretation of which companies would be captured. And right now, you are automatically part of the law if you are a QCJO company.”
For years, Narcity tried to receive a Qualified Canadian Journalism Organization (QCJO) designation, which must be determined by the Canada Revenue Agency (CRA) to qualify for journalism labor tax credits. Narcity was denied the qualification in 2021. It resubmitted in 2022, and was denied again.
“Essentially [the government] doesn’t believe that we do enough original news content to get the validation or to get the designation for our news organization, which would allow us to get the subsidies. It would allow us to get the QCJO, which would’ve included us automatically in the Online News Act,” Lapointe said.
After the denial of the QCJO, Lapointe used the rejection letter–which says that the government does not recognize it as a news organization that is engaged in the production of original news–to Meta for reconsideration. “The government does not believe that Narcity is a news organization, so why would Meta?” Two days later it regained control of its Facebook and Instagram pages.
Workarounds, circumventions, and modified hyperlinks
The ban made it difficult for news outlets and emergency officials to get information out to the public during public safety emergencies. Over the last year, officials noted that misinformation about the fires was rife on social media, encouraging the public to tune in to radio or to live blogs for accurate updates on the dozens of wildfires burning across Western Canada. As a wildfire closed in around the city of Yellowknife, Ollie Williams, the editor of Yellowknife-based news station Cabin Radio, got around the Meta ban by posting a live blog.
So, with news no longer accessible on Meta platforms, Canadians have turned to creative ways to share information that is important to them. The Media Ecosystem Observatory study revealed prior to the ban, there were about 19 posts per week with Canadian news screenshots. After the ban, this number surged to an average of 68 posts per week.
Canadians continued to share and circulate news through direct links to an X (formerly Twitter) post, which then links to the news website. They also copy and paste text.
Another way Canadians shared news on Meta platforms was by making a video with a news story in the background, talking about or responding to the events, Bridgman said. “That person, that influencer is producing engagement on the website using news content, and so they are distributing the news content.”
CJSW’s Kamis saw CJSW spelled out phonetically as a workaround. “It’s so funny. We can be on Threads because nobody’s on Threads, but this is how cynical the business model is. We’re allowed to be on Threads because no one’s on Threads.”
The future of media and news consumers at risk
Long-term ramifications for publishers are emerging, including obstacles for new launches, future growth and reaching some communities with news. According to Beers, Facebook is often the only major means of distributing information in rural, non-metropolitan areas in Canada. Many First Nations live in rural areas and Facebook acts as their source of news.
“In rural places, they’re cut off. And because Facebook purchased social license to be the place where you shared information, other alternative forms of news did not grow there,” he said. “Why would I start up a small little website in Fort Nelson or something when everyone’s on Facebook sharing news?… These places are deprived. It’ll take years, if ever, for them to come up with an alternative.”
The Observatory’s report also highlights what Bridgman called “a collective shrug,” among Canadians. Instead of actively seeking out news from reliable sources, there is a growing expectation that information will find its way to people through their existing social networks and everyday media consumption.
The behavior highlights a “news will find me” attitude. However, he said that Canadians aren’t getting their news on Facebook and Instagram any longer, so the thought is that they must be getting it somewhere else. Unfortunately, “we have very little evidence to suggest Canadians are doing that at all,” Bridgman said. “Gone are the days when you would seek out information actively, you would be an active participant in that process.”
Meta’s news ban in Canada has had a profound impact on the media landscape, though only 22% of Canadians are aware that the ban exists, according to the MEO study. It resulted in a dramatic and immediate impact on traffic, a decline in revenue, views and engagement and diminished trust, underscoring the need for new strategies to engage audiences outside of social platforms.
“I don’t think anyone looks at this and thinks this is good. It has hurt Canadian viewership of news. It has hurt Canadian news organizations. It has hurt journalism in that sense. It has hurt the ability to hold truth to power. None of that is good,” Bridgman said.
The situation Canadian news publishers find themselves in emphasizes the need for innovative business models that do not rely heavily on social media platforms for distribution and revenue.
If journalism is crucial for a healthy democracy, then we need to find ways to support its news-gathering efforts, at arms length from government, Bachynski said. Bachynski suggested implementing a taxation policy whereby digital ads and services are purchased, and a ‘journalism’ tax is applied to it, which could significantly bolster news gathering and reporting. Lapointe also mentioned the possibility of subsidizing innovations around business models.
However, as he and others have noted, the need for smart regulatory interventions becomes urgent when news organizations’ survival is tied to the whims of tech giants.
“Having this third party mediator between the content is highly risky and many have been saying that for years. And here is the proof in the pudding: relying on an external company to manage that relationship puts you at extreme vulnerability. If they decide to pull the rug out, that’s what happens,” Bridgman continued.
These days, digital media companies are all trying to figure out how to best incorporate AI into their products, services and capabilities, via partnerships or by building their own. The goal is to gain a competitive edge as they tailor AI capabilities to their audiences, subscribers and clients’ specific needs.
By leveraging proprietary Large Language Models (LLMs) digital media companies have a new tool in their toolboxes. These offerings offer differentiation and added value, enhanced audience engagement and user experience. These proprietary LLMs also set them apart from companies that are opting for licensing partnerships with other LLMs, which offer more generalized knowledge bases and draw from a wide range of sources in terms of subject matter and quality.
A growing number of digital media companies are rolling out their own LLM-based generative AI features for search and data-based purposes to enhance user experience and create fine-tuned solutions. In addition to looking at several of the offerings media companies are bringing to market, we spoke to Dow Jones, Financial Times and Outside Inc. about the generative AI tools they’ve built and explore the strategies behind them.
Media companies fuel generative AI for better solutions
Digital media companies are harnessing the power of generative AI to unlock the full potential of their own – sometimes vast amounts – of proprietary information. These new products allow them to offer valuable, personalized, and accessible content to their audiences, subscribers, customers and clients.
Take for example, Bloomberg, which released a research paper in March detailing the development of its new large-scale generative AI model called BloombergGPT. The LLM was trained on a wide range of financial data to assist Bloomberg in improving existing financial natural language processing (NLP) tasks, such as sentiment analysis, named entity recognition, news classification, and question answering, among others. In addition, the tool will help Bloomberg customers organize the vast quantities of data available on the Bloomberg Terminal in ways that suit their specific needs.
Launched in beta June 4, Fortune partnered with Accenture to create a generative AI product called Fortune Analytics. The tool delivers ChatGPT-style responses based on 20 years of financial data from the Fortune 500 and Global 500 lists, as well as related articles, and helps customers build graphic visualizations.
Generative AI helps customers speed up processes
A deeper discussion of how digital media companies are using AI provides insights to help others understand the potential to leverage the technology for their own needs. Dow Jones, for example uses Generative AI for a platform that helps customers meet compliance requirements.
Dow Jones Risk Compliance is a global provider of risk and compliance solutions across banks and corporations which helps organizations perform checks on their counterparties. They do that from the perspective of complying with anti-money laundering regulation, anti-corruption regulation, looking to also mitigate supply chain risk and reputational issues. Dow Jones Risk Compliance provides tools that allow customers to search data sets and help manage regulatory and reputational risk.
In April, Dow Jones Risk & Compliance launched an AI-powered research platform for clients that enables organizations to build an investigative due diligence report covering multiple sources in as little as five minutes. Called Dow Jones Integrity Check, the research platform is a fully automated solution that goes beyond screening to identify risks and red flags from thousands of data sources.
The planning for Dow Jones Integrity Check goes back a few years, as the company sought to provide its customers with a quicker way to do due diligence on their counterparties, Joel Lange, executive Vice President and General Manager, Risk and Research at Dow Jones explained.
Lange said that Dow Jones effectively built a platform which automatically creates a report for customers on a person or company, using technology from AI firm Xapien. It brings together Dow Jones’ data that is plugged into other data sets, corporate registrar information, and wider web content. It then leverages the platform’s Generative AI capability to produce a piece of analysis or a report.
Dow Jones Risk & Compliance customers use their technology to make critical, often complex, business decisions. Often the data collection process can be incredibly time consuming, taking days if not weeks.
The new tool “provides investigations, teams, banks and corporations with initial due diligence. Essentially it’s a starting point for them to conduct their due diligence, effectively automating a lot of that data collection process,” according to Lange.
Lange points out that the compliance field is always in need of increased efficiency. However, it carries with it great risk to reputation. Dow Jones Integrity Check was designed to reshape compliance workflows, creating an additional layer of investigation that can be deployed at scale. “What we’re doing here is enabling them to more rapidly and efficiently aggregate, consolidate, and bring information to the fore, which they can then analyze and then take that investigation further to finalize an outcome,” Lange said.
Regardless of the quality of the generated results, most experts believe that it is important to have a human in the loop in order to maintain content accuracy, mitigate bias, and enhance the credibility of the content. Lange also said that it’s critical to have “that human in the loop to evaluate the information and then to make a decision in relation to the action that the customer wants to take.”
In recent months, digital media companies have been launching their own generative AI tools that allow users to ask questions in natural language and receive accurate and relevant results.
The Associated Press created Merlin, an AI-generated search tool that makes searching the AP archive more accurate. “Merlin pinpoints key moments in our videos to exact second and can be used for older archive material that lacks modern keywords or metadata,” explained AP Editor in Chief Julie Pace at The International Journalism Festival in Perugia in April.
Outside’s Scout: AI search with useful results
Chatbots have become a popular form of search. Originally pre-programmed and only able to answer select questions included in their programming, chatbots have evolved and increased engagement by providing a conversational interface. Used for everything from organizing schedules and news updates to customer service inquiries, Generative AI-based chatbots assist users in finding information more efficiently across a wide range of industries.
Much like The Guardian, The Washington Post, The New York Times and other digital media organizations that blocked OpenAI from using their content to power artificial intelligence, Outside CEO Robin Thurston explained that Outside Inc. wasn’t going to let third parties scrape their platforms to train LLM models.
Instead, they looked at leveraging their own content and data. “We had a lot of proprietary content that we felt was not easily accessible. It’s almost what I’d call the front page problem, which is you put something on the front page and then it kind of disappears into the ether,” Thurston said.
“We asked ourselves: How do we create something leveraging all this proprietary data? How do we leverage that in a way that really brings value to our user?” Thurston said. The answer was Scout, Outside Inc.’s AI search assistant. Scout is a custom-developed chatbot.
The company could see that generative AI offered a way to make that content accessible and even more useful to its readers. Outside had a lot of evergreen content that wasn’t adding value once it left the front page. Their brands inspire and inform audiences about outdoor adventures, new destinations and gear – a lot of which is evergreen and proprietary content that still had value if it could easily be surfaced by its audience. The chat interface allows their content to continue to be accessible to readers after it is no longer front and center on the website.
Scout gives users a summary answer to their question, leveraging Outside Inc’s proprietary data, and surfaces articles that it references. “It’s just a much more advanced search mechanism than our old tool was. Not only does it summarize, but it then returns the things that are most relevant,” he explained.
Additionally, Outside Inc’s old search function worked by each individual brand. Scout searches across the 20+ properties owned by the parent company which include Backpacker, Climbing, SKI Magazine, and Yoga Journal, among others. Scout brings all of the results together, from the 20+ different Outside brands, from the best camping destinations, to the best trails, outdoor activities for the family, gear, equipment and food all in one result.
One aspect that sets Outside Inc.’s model apart is their customer base, which differs from general news media customers. Outside’s customers engage in a different type of interaction, not just a quick transactional skim of a news story. “We have a bit of a different relationship in that they’re not only getting inspiration from us, which trip should I take? What gear should I buy? But then because of our portfolio, they’re kind of looking at what’s next,” Thurston said.
It was important to Thurston to use the LLM in a number of different ways, so Outside Inc launched a local newsletter initiative with the help of AI. “On Monday mornings we do a local running, cycling and outdoor newsletter that goes to people that sign up for it, and it uses that same LLM to pick what types of routes and content for that local newsletter that we’re now delivering in 64,000 ZIP codes in the U.S.”
Thurston said they had a team working on Scout and it took about six months. “Luckily, we had already built a lot of infrastructure in preparation for this in terms of how we were going to leverage our data. Even for something like traditional search, we were building a backend so that we could do that across the board. But this is obviously a much more complicated model that allows us to do it in a completely new way,” he said.
Connecting AI search to a real subscriber need
In late March, The Financial Times released its first generative AI feature for subscribers called Ask FT. Like Scout, the chat-based search tool allows users to ask any question and receive a response using FT content published over the last two decades. The feature is currently available to approximately 500 FT Professional subscribers. It is powered by the FT’s own internal search capabilities, combined with a third-party LLM.
The tool is designed to help users understand complicated issues or topics, like Ireland’s offshore energy policy, rather than just searching for specific information. Ask FT searches through Financial Times (FT) content, generates a summary and cites the sources.
“It works particularly well for people who are trying to understand quite complex issues that might have been going on over time or have lots of different elements,” explained Lindsey Jayne, the chief product officer of the Financial Times.
Jayne explained that they spend a lot of time understanding why people choose the FT and how they use it. People read the FT to understand the world around them, to have a deep background knowledge of emerging events and affairs. “With any kind of technology, it’s always important to look at how technology is evolving to see what it can do. But I think it’s really important to connect that back to a real need that your customers have, something they’re trying to get done. Otherwise it’s just tech for the sake of tech and people might play with it, but not stick with it,” she said.
Trusted sources and GenAI attribution
Solutions like those from Dow Jones, FT and Outside Inc. highlight the power of a brand with a trusted audience relationship to create deep, authentic relationships built on reliability and credibility. Trusted media brands are considered authoritative because their content is based on credible sources and facts, which ensures accuracy.
Currently, generative AI has demonstrated low accuracy and poses challenges to sourcing and attribution. Attribution is a central feature for digital media companies who roll out their own generative AI solutions. For Dow Jones compliance customers, attribution is critical to customers, to know if they’re going to make a decision based on information that is available in the media, according to Lange.
“They need to have that attributed to within the solution so that if it’s flowing into their audit trails or they have to present that in a court of law, or if they would need to present it to our internal audit, the attribution is really key. (Attribution) is going to be critical for a lot of the solutions that will come to market,” he said. “The attribution has to be there in order to rely on it for a compliance use case or really any other use case. You really need to know where that fact or that piece of information or data actually came from and be able to source it back to the underlying article.”
The Financial Times’ generative AI tool also offers attribution to FT articles in all of its answers. Ask FT pulls together lots of different source material, generates an answer, and attributes it to various FT articles. “What we ask the large language model to do is to read those segments of the articles and to turn them into a summary that explains the things you need to know and then to also cite them so that you have the opportunity to check it,” Jayne said.
They also ask the FT model to infer from people’s questions when it should be searching from. “Maybe you’re really interested in what’s happened in the last year or so, and we also get the model to reread the answer, reread all of the segments and check that, as kind of a guard against hallucination. You can never get rid of hallucination totally, but you can do lots to mitigate it.”
The Financial Times is also asking for feedback from the subscribers using the tool. “We’re literally reading all of the feedback to help understand what kinds of questions work, where it falls down, where it doesn’t, and who’s using it, why and when.”
Leaning into media strengths and adding a superpower
Generative AI seems to have created unlimited opportunities and also considerable challenges, questions and concerns. However it is clear that an asset many media companies possess is a deep reservoir of quality content and it is good for business to extract the most value from the investment in its creation. Leveraging their own content to train and program generative AI tools that serve readers seems like a very promising application.
In fact, generative AI can give trustworthy sources a bit of a super power. Jayne from the FT offered the example of scientists using the technology to read through hundreds of thousands of research papers and find patterns in a process that would otherwise take years to read in an effort to make important connections.
While scraped-content LLMs pose risks to authenticity, accuracy and attribution, proprietary learning models offer a promising alternative.
As Jayne put it, “The media has “an opportunity to harness what AI could mean for the user experience, what it could mean for journalism, in a way that’s very thoughtful, very clear and in line with our values and principles.” At the same time, she cautions that we shouldn’t be “getting overly excited because it’s not the answer to everything – even though we can’t escape the buzz at the moment.”
We are seeing many efforts bump up against the limits of what generative AI is able to do right now. However, media companies can avoid some of generative AI’s current pitfalls by employing the technology’s powerful language prediction, data processing and summarization capabilities while leaning into their own strengths of authenticity and accuracy.
In the age of artificial intelligence, it could be argued that the calculus of content is changing.
Since the advent of publishing metrics, the goal has always been more: more page views, clicks, keywords and SEO. And while AI can automate various aspects of content creation and production to save digital media companies time and resources, the convenience of the technology has also allowed for a firehose of low quality content to proliferate.
But, amid the sharp increase of these junk content farms, is there a new opportunity for quality journalism to quietly reclaim its place at the fore? It’s certainly on the minds of the leadership at The Atlantic.
In April, The Atlantic announced it had reached 1 million subscribers and become profitable, by investing in areas where the company had “fairly high confidence of good returns,” according to CEO Nicholas Thompson. The 167-year-old publication currently boasts financial stability and is well-positioned to think about where it is headed as it approaches its 200th birthday.
The Atlantic’s one-million milestone is just the foundation for further growth. In a recent memo to staff, Thompson and Editor in Chief Jeffrey Goldberg wrote:
“The key to continued success is to be constructively dissatisfied with the present, and so both of us believe very strongly that our 1 million subscriptions represent merely the foundation of future excellence and growth.”
Goldberg says that he would like The Atlantic to double, then quadruple its current size, saying the company needs to figure out how to reach larger audiences around the world. “I want to set a course as The Atlantic heads towards its bicentennial in 33 years. Now is the time for us to decide this is where we want The Atlantic to go and this is how we’re going to get there,” he said.
In terms of excellence, The Atlantic’s awards speak volumes to the quality of the journalism it produces. For the third year in a row, The Atlantic was awarded General Excellence for a News, Sports, and Entertainment publication at the 2024 National Magazine Awards. No one else has done that in this century, Goldberg remarks.
“So, we have the recognition of our industry that we’re doing something right, and I feel like this is the year when we need to really focus on: what are the next large steps we take?” Goldberg said. “Because we’re in a very good spot. But I don’t want to spend the next five years defending the hill that we’re on. I want to move to some other mountain entirely.”
And, that means not just defending their reputation for journalistic excellence, or growing iteratively. It means figuring out how to reach enormous audiences around the world – and getting a whole lot of them to subscribe.
Subscription strategy with an editorial focus
Prior to the pandemic, Atlantic Chair David Bradley and Laurene Powell Jobs decided they should move into the digital subscription space. They’d had a lot of success through scale – growing web traffic and advertising, Goldberg said. The company hired Alexandra Hardiman, (currently New York Times’ Chief Product Officer), as Chief Business & Product Officer in 2018-2019, to build The Atlantic’s digital subscription model.
“We launched basically six months before the pandemic started when advertising collapsed. And we did very well in those early months. There was a lot of demand,” Goldberg recalled.
The 2019 metered model offered three annual subscription plans for readers: digital, print and digital and a premium tier, which offered exclusive access to podcasts, product discounts and priority access to events among other perks. The Atlantic earned 300,000 new subscriptions in the 12 months that followed.
The following year, the pandemic impacted the publication’s in-person events and advertising, forcing layoffs of 17% of its staff, and losses in the millions.
Now, overall revenue is up more than 10% year over year. The company says advertising booked year-to-date is also up 33% year over year. And, subscriptions to The Atlantic have increased by double-digit percentages in each of the past four years. In fact, they’ve surged 14% in the past year alone.
By 2023, The Atlantic was back on the path to profitability. According to Axios, The Atlantic adjusted its paywall to be more flexible for subscribers and was working to add new revenue streams. Then, roughly 60% of its revenue came from subscriptions, which included print magazines and digital subscriptions through Apple News.
Flexible paywalls meet journalistic excellence
The strategy was to have the best, smartest, most dynamic, flexible subscription, acquisition and retention strategies, Goldberg said. “We’ve always believed that you can have the best systems in the world for acquiring people easily, but if you don’t have a quality product to sell them, they’re not going to come, they’re not going to stay.”
“We pivoted, I would say, to a total quality model on the web. We were doing good stuff on the web for years. We had a large team of young reporters doing news analysis and quick summaries and that sort of thing. But I’ve always believed that the aspect of The Atlantic that differentiated us from everyone else was a commitment to having the highest standards and producing the most complicated, interesting, aesthetically-pleasing, well-written journalism. I think that strategy has borne fruit,” Goldberg said.
The Atlantic focused on editorial excellence, publishing stories that exemplified depth and range and drove news cycles. It recruited high-profile writers including New York Times’ Jennifer Senior and Caitlin Dickerson, who won Pulitzers in 2022 for Feature Writing and Explanatory Journalism, respectively.
“My goal here is to build the greatest writers collective in the English language. We’re halfway there,” Goldberg said. “I don’t need the biggest one. I just need the best one. There are tremendous numbers of readers of English, who want access to our writers, and so as long as there’s an audience for quality journalism, quality non-fiction, we will be okay.”
Reaching new subscribers with newsletters
In addition to a flexible digital subscription strategy and editorial excellence, The Atlantic invested in new newsletters for subscribers in 2021, bringing nine newsletter writers into the fold. Newsletters help reach different audiences, build loyalty and repetition. And as Goldberg pointed out, The Atlantic is launching new ones all of the time. Thus far, the strategy appears to be a moderate success.
“One of the best things to happen out of that is we found more great staff writers, Yair Rosenberg, Xochitl Gonzalez and Charlie Warzel, just to name three and so, it ultimately brought their following,” Goldberg said. “They’re integrated into our writers collective in a way that’s great for our readers and great for our journalism.”
High-quality content reckons with AI
As digital media companies reckon with the changes artificial intelligence brings, deciding on how to adapt or adopt, it’s becoming clear that high-quality journalism retains immense value in the AI era. It offers authenticity, context, and deep analysis that AI-generated content lacks. It provides meaningful insights, informs people and counters misinformation.
Despite the one million subscriber milestone, Goldberg isn’t ready to relax. “It’s not like a breath out. We’re not breathing easy because you’ve got to run scared in this business,” he said. “But those three things, the subscription health, financial health, journalism health and recognition, give us a great place to have meetings where we can actually think through, alright, what are we going to do with The Atlantic on its approach to its 200th birthday.”
“Because it is a very unstable industry, obviously, and I worry about small mistakes or small missed opportunities snowballing over the years. I worry about missing the opportunity to do something newer and bigger.”
Few of The Atlantic’s contemporaries are left. As Goldberg points out, many venerable magazines that came after The Atlantic – like Collier’s and The Saturday Evening Post – have disappeared.
“So it’s kind of a miracle that The Atlantic has made it through the beginning of the internet age successfully. It survived the Great Depression and the Civil War and World War II. And, so we really have to focus on what is it that made it survive? And what do we do to increase its chances of surviving and flourishing into the next phase?”
Advertising and subscriptions are popular podcasting revenue streams. As with all things digital media, revenue diversification is the name of the game. So, it is interesting to see media companies exploring alternative and creative revenue streams for podcasting.
Deloitte predicts that a growing number of consumers worldwide will engage with audio entertainment overall in 2024—bringing the number of monthly average podcast listeners to more than 1.7 billion. Fueled by this demand, podcast advertising is also on the rise. In fact, it outpaces growth trends for the internet ad market, according to IAB’s U.S. Podcast Advertising Revenue Study 2023. Podcast advertising revenues have grown over 115% from 2020 to 2022. And ad revenues grew 26% year-over-year to $1.8 billion.
So, it’s no surprise that for many publishers, networks and individual creators, advertising is still the primary revenue driver. Jenna Weiss-Berman is EVP, Head of Podcasts at Audacy, Inc. and Co-Founder, Pineapple Street Media (an Audacy Company). She believes that, while it is important to explore new revenue opportunities, advertising is too important to neglect.
“We all want to look into alternative models,” Weiss-Berman said. “But I think we are also very focused on the one that we know is working and how do we make that one work better? I think because we are seeing a lot of annual and quarterly growth, we want to make sure that that’s really the priority – focusing on the thing that we know works best.”
Ray Chao, SVP and general manager of audio and digital video at Vox Media, also believes that advertising is still a strong revenue model for podcasts. He points to a combination of new advertisers entering the podcast space as well as existing advertisers allocating increasing budgets to podcasting. “I think what that reflects is a real understanding of the fact that podcasts are one of the most engaging media channels,” he said.
Streaming ad innovations for podcasters
And, the research certainly backs up the level of engagement. Podcast listeners in the U.S. are expected to spend an average of 54 minutes per day tuned into programming in 2024, according to EMarketer’s February 2024 forecast. According to eMarketer, in 2024 digital audio will make up a fifth of all time spent with digital media in the U.S. and in 2023, almost 75% of U.S. internet users listened to digital audio.
“This growth comes as no surprise to us,” said Emma Vaughn, global head of advertising business development and partnerships at Spotify. Spotify’s advertising revenue climbed 12% to €501 million last year—an all-time high, thanks to double-digit growth across both music and podcasts. With this kind of revenue opportunity on the table, Spotify is focused on evolving the advertising technology for podcasting.
“We’re continuously advancing the ad products we offer,” Vaughn said. The company just launched a music agency called AUX in February. “We’ll use our expertise to counsel brands about how best to use music to enrich their campaigns and leverage our relationships with emerging artists to help them reach new audiences.”
The company has also been “been modernizing podcast ads in format and in experience through CTA Cards, an interactive, multi-way experience that transformed podcast ads from something that could only be heard, into an experience that you can also see — and, most importantly, click,” she said.
Vaughn also emphasized the need for measurement to mature in the podcast business, citing the global shift in consumption from downloads to streaming which has opened the door for advances in more data-driven podcast advertising measurement. “When we launched Streaming Ad Insertion (SAI), we introduced impression-level insight to podcast advertising for the very first time,” she explained. “Rather than relying on download data, advertisers now had access to standard digital reporting like impressions, unique reach, and frequency, along with unique audience insights and creative performance metrics powered by Spotify’s first-party data.”
SAI is a podcast ad technology available on Spotify that incorporates digital planning, reporting, and measurement for podcast advertisers – based on confirmed ad impressions. It was designed to “deliver the intimacy and quality of traditional podcast ads with the precision and transparency expected of digital marketing” according to Vaughn.
Paid podcast subscriptions
However, Vaughn points out that at Spotify, subscriptions are a critical piece of their business. “In our last earnings, subscription revenue increased 17% to €3.17 billion.”
And, while most podcasts are ad-supported and free to listeners, some publishers are now offering exclusive podcast content only paid subscribers can access. For example, late last year The Economist launched an audio subscription ($4.90 US per month) that put the majority of its shows behind a paywall. The Economist’s entire collection of podcasts is exclusively available to Economist Podcasts+ subscribers.
At Fox News Audio, the breadth of podcast content across the Fox Audio Network allows the company to attract a broader audience with diverse interests, explained William Sanchez, executive director of digital & business development at Fox News Audio and Fox Audio Network. Fox News is not new to podcasting or subscriptions.
The company has been in the business for 10 years, and Sanchez says that “when we started our first venture into podcasts with our talk radio shows as premium podcasts. So they’ve always been a part of our strategy as far as subscription goes, and now is inclusive of Apple podcast subscriptions and ad-free listening on Amazon Prime for Prime members. Subscription continues to be, it was always a part of our plan or strategy, and it used to be a part of our roadmap.”
At Vox Media, while the company’s business model is primarily advertising-driven and their main focus is continuing to grow advertising. However, it’s also a priority to continue to diversify their revenue model, explained Chao. “We’ve done that through subscriptions and consumer revenue,” Chao said. “We’ve been operating Cafe Insider, which is a subscription business for several of our shows, including Stay Tuned with Preet. And, we recently launched Criminal+, which is a subscription product for Criminal. We’re excited about consumer revenue as a space to continue to grow and diversify our revenue model.”
Their strategy starts by looking at the show and thinking about its host, its team, the audience, and figuring out if there is an opportunity for a paid consumer product that allows fans of the show to engage more deeply with the content, Chao said. “So, for fans of Preet Bharara who are listening to Stay Tuned with Preet, they are excited about more content in the intersection of policy, law and politics. And, we deliver that each week through Cafe Insider.”
Video audiences expand podcast audience and revenue
While podcasts used to be the medium of choice for multi-taskers, publishers are now filming their podcasts and distributing them on visual platforms with big built-in audiences like YouTube and TikTok. In case media companies worry that offering full length content might cannibalize an audio audience, there’s certainly a wealth of highlights to be clipped for social and promotional purposes, and a clear opportunity to expand podcast monetization.
The fact is that people actually enjoy watching video podcasts almost as much as listening to them. A February YouGov poll suggested 33% of respondents favored listening to audio podcasts, while 25% preferred watching video podcasts. The visual appeal of video podcasts was one of the main reasons for 51% of respondents, while 29% preferred the variety of content and ability to share on social platforms.
And, as users increasingly consume podcasts in both audio and video formats, leveraging YouTube for consumption, digital media companies are slowly expanding upon existing ad models to make video part of their go-to-market plans. Video-enabled podcasts represented just under 10% of revenues in 2023, according to IAB’s most recent U.S. Podcast Advertising Revenue Study. Which makes it a prime opportunity for advertisers to engage in cross-channel buys.
Last year, Slate announced a partnership with YouTube to bring Slate’s network of podcasts, including favorites like Slow Burn, Political Gabfest, What Next and Amicus, to the platform.
“We see this as a real opportunity to build scale and reach a new, untapped audience on YouTube, which has become the world’s most-used podcast platform,” wrote Charlie Kammerer, Slate’s president and CRO, in a release.
After introducing video podcasts in 2020, Spotify announced in June last year that they had more than 100,000 video podcasts on their platform, adding that the U.S., Brazil, Mexico, the U.K., and Germany published the most video podcasts.
Fox News Audio has success with video and its podcasts, according to Sanchez. “Our audio division lives among a video first world, so we get to tap into a lot of the resources and the expertise from the video side of the business that allows us to bring the audio-first business into the video world as well and capitalize our platforms like YouTube.”
So, while Fox News Audio has shows that come from the video side, they’re findinging that they have a much larger catalog that is audio-based. “We’re starting to move forward with bringing those to our YouTube channel. So, that’s another revenue stream opportunity for us as we all try to figure out what video role plays in podcasting.”
As Chao sees it, a podcast is just a direct to consumer (D2C) show that has been delivered through RSS onto platforms like Apple and Spotify. However, the ecosystem for podcasting continues to expand into new content categories, new creators and talent, and as consumer behavior continues to evolve and there are flourishing video platforms like YouTube and TikTok out there.
“We’ve seen a lot of success with these direct to consumer shows, building an audience and a business on platforms outside of just traditional RSS feed podcast listening platforms,” said Chao. “So, if you think about podcasts more broadly than just an audio show, it’s actually a direct to consumer show that can be audio, can be video. That is really where we see continued expansion in the podcast industry for both audiences and advertisers.”
Podcast merchandise revenue: swag, gifts and more
Podcast networks and digital publishers are now hawking merchandise – from clothing lines to bags, water bottles and more. Fans, eager to show their allegiance to their favorite podcasts through t-shirts, mugs, or hoodies, discuss which podcasts have the best merchandise.
Fox News Audio offers some podcast-related merchandise as well. While it’s not a large part of their business, there is representation in the Fox shop, Sanchez said.
American podcast network Wondery, home to podcasts Morbid, Smartless, Redhanded, MrBallenand other popular shows, has a merchandise shop that offers apparel, accessories, drinkware and home and office.
Podcast revenue forecast
“The era of podcasting that we’re moving into now is that we need it to be a sustainable business and not just something that people are trying to get rich quick off of,” according to Weiss-Berman. “I think that it’s fun figuring out how a thing can be sustainable and a good long-term business. There’s a lot of opportunity for growth that we’re seeing and it’s actually not very complicated to make it a profitable business when that’s the focus.”
While 2023 was a challenging year for podcasters, with layoffs at NPR, Spotify, Stitcher and Amazon Music and lucrative deals with celebrities falling apart, the industry also saw growth in listeners and ad revenue. While it is heartening to see stalwart revenue models–advertising and subscription supported–holding strong for podcasting, diversification builds strong media businesses. Publisher investment in new revenue streams like video and merchandise comes from a desire to drive expansion, growth and sustainability.
As Vox’s Chao put it: “As we think about continuing to grow this business and to continue to help it flourish, these additional revenue streams or diversifying our revenue model through expansion into subscriptions and consumer revenue and other types of revenue streams is a core component of how we’re going to continue to grow this business going forward.”
