Articles by: "Kevin Rehberg, Vice President, Client Development – Alliance for Audited Media"

As the media’s use of artificial intelligence grows, so do questions from audiences and advertisers about how they’re using it. While media companies experiment and find ways to integrate AI tools into their businesses, there’s also a need for increased transparency and disclosure around their use of AI.

Here are several ways media companies are leveraging AI in their operations and practical tips for enhancing transparency and fostering open communication with audiences.

How media companies are implementing AI

If AI could be summed up in one word in the past year, it would be experimentation. Many media companies have begun to add AI tools to their workflows to accomplish tasks such as brainstorming headlines, transcribing interviews, summarizing research or analyzing data. These use-cases help media companies improve workflows and increase efficiency.

Jennifer Bertetto, president and CEO of Trib Total Media and 535media, said her company employs AI-driven solutions to streamline reporting processes and surface content for its audience.

“AI is instrumental in personalizing content recommendations, refining advertising strategies and driving overall business efficiency,” Bertetto said. “By harnessing AI’s capabilities, we are committed to delivering superior content, fostering deeper audience engagement and ensuring sustainable growth.”

Best practices for disclosing AI use

A recent survey by Trusting News and Online News Association (ONA) revealed that nearly 94% of news consumers want newsrooms to disclose how they use AI. And more than half seek to understand what tools were implemented and how they were used during the reporting process.

To help media companies determine how to communicate this to their audiences, Poynter developed a guide to creating an AI ethics policy, which includes recommendations such as forming an AI committee and including representatives from all departments to weigh in on issues and set policies.

Aaron Kotarek, who was recently senior vice president of audience and operations for the Honolulu Star-Advertiser/Oahu Publications, Inc., and is now general manager and chief operating officer at The Spokesman-Review, led the Star-Advertiser’s AI task force. “We took great pride at OPI as it pertained to accuracy and credibility. We did not want to risk our reputation in the communities we serve by executing something rash,” Kotarek said.

Poynter also recommends dividing AI decision-making into three categories — audience-facing, business uses and back-end reporting. They also suggest that media organizations develop standards around each of these areas.

Industry resources for media companies

In addition to Poynter’s starter kit for newsrooms, there are additional resources media companies can turn to for guidance.

Trusting News, an organization to help journalists earn audience trust, created an AI Trust Kit for newsrooms to help media companies establish internal guidelines and public-facing policies for governing AI practices. The kit also offers tips on what to disclose including why journalists use AI, how they use it and how human oversight has been integrated into the editorial process.

The National Institute of Standards and Technology (NIST) created an AI Risk Management Framework to improve the trustworthiness of AI. The U.S. Department of Commerce developed this document to help companies think about, communicate, measure and monitor AI’s potential risks and positive impacts.

The Alliance for Audited Media is also working with publisher clients to help them navigate AI concerns and understand and implement industry best practices, drawing on experience certifying media companies in privacy and data protection programs and journalism ethics.

As AI continues to shape the media landscape, its responsible implementation and transparent disclosure is key to maintaining audience trust. By embracing best practices and leveraging industry resources, media companies can continue to innovate while upholding their commitment to quality and accountability.

Innovation | Perspectives | Supporter Spotlight

3 tools to make media companies stand out in the age of AI

March 20, 2023 Kevin Rehberg, Vice President, Client Development – Alliance for Audited Media

Artificial intelligence chatbots such as OpenAI’s ChatGPT and Microsoft’s BingGPT have generated lots of buzz lately, much of it centered around how helpful or harmful these tools might be for media companies. While the helpful camp welcomes the benefits of assisting reporters with initial research, aggregating data and performing basic copy editing, others argue that these programs may also accelerate the spread of misinformation and fake news sites, which can compete with legitimate publishers for readers’ attention and advertising.

While newsrooms have been using AI tools to some degree for years, the technology has recently become more advanced and widespread, making it harder to determine whether copy was written by a human or generated by a bot. Publishers committed to industry standards and best practices need to find ways to stand out from lesser quality websites.

Oddly enough, even ChatGPT agrees that publishers need to find ways to distinguish themselves from non-human creators. When asked how digital publishers can stand out from AI-generated content, it responded:

“As AI-generated content becomes increasingly prevalent, media companies must find ways to differentiate themselves in the marketplace. While AI can produce articles and other types of content at a rapid pace, there are certain aspects of digital publishing that only a human touch can provide.”

What are those human touches? Fact-checking, vetting sources and objective reporting, to name a few. As a legitimate digital publisher, you already do these things, but do you get credit for them?

Here are three initiatives designed to help media companies stand apart from artificial content and get more credit for their good work.

Showcase your commitment to journalism standards and ethics

As more publishers use AI tools, questions may arise about whether they followed journalistic standards during the content creation process.

The Journalism Trust Initiative (JTI) is a certification program launched by Reporters Without Borders (RSF) to help quality media outlets demonstrate their commitment to creating transparent, ethical journalism. The program includes certification against the JTI Standard – a set of transparency standards developed to measure media outlets’ production of trustworthy content.

“It’s really important to be able to distinguish between trusted news and misinformation,” said Beth Potter, Ph.D., U.S. regional manager for JTI. “We as journalists are not marketing ourselves well enough. We need to stand up as a group and let the world know we have rules and policies in place that lesser-quality content providers do not have.”

The three-step JTI certification process includes a self-assessment, a public disclosure in the form of a Transparency Report and a third-party audit to confirm the organization’s adherence to the JTI Standard. Once certified, media outlets can display the JTI mark to let advertisers know they are committed to following industry standards that lead to the creation of trustworthy content.

Recognize your industry connections and memberships

Quality media outlets often take an active role in the industry and participate in groups like Digital Content Next and the Alliance for Audited Media to promote quality media and industry standards.

Trust.txt is a framework developed by non-profit organization JournalList to help organizations earn recognition for their relationships with trusted industry organizations and associations. Similar to Ads.txt, publishers place a Trust.txt file on their websites that publicly lists industry memberships, owned domains and social media accounts. Industry organizations are also encouraged to create a file listing their members. The goal is to help search engines recognize the legitimate connections between media companies and industry associations.

“A great many of us depend on unreliable mechanisms to find reliable news online,” said Mark Stencel, executive director of JournalList. “Tech companies shouldn’t be the ones who decide what is and what is not ‘news.’ Journalism organizations should. JournalList’s Trust.txt files are a simple and efficient way to make that possible.”

Get certified for creating balanced, reliable content

Brand safety has been an increasing concern as advertisers find ways to prevent their ads from running on websites that don’t align with their brand values or create misleading content.

Ad Fontes Media developed its interactive Media Bias Chart® in response to increases in misinformation and media polarization. By creating a methodology to rate sources for reliability and media bias, Ad Fontes offers a tool to help advertisers and consumers find trusted sources of news and information. The group rates everything from websites, linear and connected TV, YouTube channels, podcasts and newsletters.

“When people say they don’t trust media, they lump quality news sources with those that create questionable content,” said Vanessa Otero, CEO and founder of Ad Fontes Media. “We want to elevate quality publishers and showcase their good work and commitment to transparency.”

The Alliance for Audited Media recently partnered with Ad Fontes to create a custom media bias chart featuring AAM-audited publishers, which Otero says tend to be among the most reliable, trustworthy sources Ad Fontes rates across the news landscape.

“Advertisers use the chart to create inclusion lists based on media bias and reliability ratings,” Otero said. “The chart helps advertisers support reliable and trustworthy sources while safeguarding their brands.”

Advertisers also use our AAM Audited Domain List, a list of digital publishers who have completed the AAM Digital Publisher Audit, to create inclusion lists and direct more of their investment toward publishers that have been vetted by a third party. The audit gives media companies another tool to help stand out for their adherence to industry standards and best practices for sourcing website traffic.

The future is in publishers’ hands

As AI technology continues to evolve, it will only continue to become more challenging to separate content written by humans from articles generated by chatbots. It’s even challenging for ChatGPT, which when asked about its ability to identify AI-generated content, replied:

“As an AI language model, I can analyze and generate text, but I cannot determine with 100% accuracy whether a piece of content was generated by a human or an AI system.”

While publishers may not be able to prevent the spread of chatbot-sourced content, they can control how their websites appear to readers, advertisers, and search engines. By implementing industry initiatives, publishers can do something the chatbots can’t do: prove that they are committed to standards and best practices and provide transparency about how their content was sourced and developed.

Perspectives | Supporter Spotlight

Humans vs. Bots: It pays to find out who is really looking at your content

August 31, 2022 Kevin Rehberg, Vice President, Client Development – Alliance for Audited Media

Digital publishers work hard to create quality content that engages audiences and keeps them coming back for more. When an article or video receives a surge in views, that means mission accomplished, right? Not so fast.

Since invalid traffic (IVT) makes up for nearly 40% of all website traffic, it’s likely that some of these viewers are non-human, or “bots.” And, unfortunately, when bot traffic is included in analytics reports, it can skew data and give an inaccurate look at content performance.

Here are steps publishers can take to achieve more accurate website data, which allows them to make more informed editorial and marketing decisions.

Identify traffic sources and spikes

To understand how content and promotions are truly performing, it’s important to identify and remove invalid traffic from data reports. A first step is to determine traffic sources. If a visitor comes from an unusual website or unfamiliar referral source, it can be an indication that the traffic is not human. This includes traffic from locations beyond the typical readership area. Or it might be a slew of visitors from one city, which could be the site of a datacenter.

Another indicator could be the time of day that visitors view content. If many arrive on the site at an unusual time such as in the middle of the night, it could indicate that the traffic is robotic. Other non-human behavior such as spending zero seconds on a page or achieving a 100% bounce rate are also signs the traffic might be robotic.

Analyzing traffic spikes might also uncover increases in legitimate visitors, which offers insight into why content performed well. For example, an increase in an article’s views could indicate that it was picked up by another news source, shared on a popular website, or received a greater attention on social media. Identifying traffic sources and patterns can reveal what articles are truly performing well. And, of course, that is likely to influence future content decisions.

Filter bot traffic

Not all bot traffic is malicious. Search engines send bots to crawl websites to learn more about them, which helps inform search results. However, bots also visit websites for harmful reasons including scraping content to create fake sites, inserting malware to steal user data and injecting spam. Publishers can also unintentionally invite bots to their website by purchasing website traffic.

Many analytics platforms have built-in tools to filter bot traffic. Since these tools often must be activated by a site administrator to begin working, it’s helpful to check to make sure the filter is being applied to analytics reports.

While automatic known bot filtering tools are a great start to getting a more accurate look at website data, they might not detect all bots. Publishers should identify traffic patterns of good and bad user behavior to identify unknown robots that are not detected by the analytic partner and create custom filters based on the characteristics of those bots detected. Some of these characteristics may include the location the traffic came from, screen resolution, service provider, time on page and bounce rate.

Identifying traffic patterns and creating custom filters to remove bots from data ensures that publishers are getting the most accurate look possible at their website visitors.

Check site tags

Analytics providers require that publishers install a tag on their website to allow them to collect data. If a tag is inadvertently installed more than once, data analytics will be inflated. Or, if installed incorrectly, it might produce incomplete data. Publishers should check tag containers to ensure there is only one tag for each tracking suite so that data isn’t included twice in reports.

Get a third-party evaluation

Sometimes it helps to get input from a third party that specializes in analyzing website traffic. A digital publisher audit is great way to get an expert’s opinion about site traffic, tagging and creating custom filters. Website auditors keep up with industry trends and platform changes, which saves publishers time and the resources needed to identify new bots or create custom filters.

By taking the steps above, publishers can better identify high performing content, discovering promotions that drive real traffic, and making more informed business decisions. They can also take steps to reduce bot traffic and avoid common missteps that occur when it goes undetected.

Advertising | Perspectives | Revenue

The next frontier in digital ad fraud: local markets

October 29, 2018 Kevin Rehberg, Vice President, Client Development – Alliance for Audited Media

Digital advertising fraud continues to be an industry-wide problem. Estimates of dollars lost to fraud range from $6 billion to $16 billion annually. Much of the conversation has been on preventing fraud in open ad exchanges on a national level. But digital ad fraud can affect anyone who buys and sells digital advertising, including local and national publishers and advertisers.

With local markets becoming a target, the ad fraud crisis might seem to be spiraling out of control. But there is hope. The war on ad fraud is winnable. National and local marketers have the power to steer their budgets toward legitimate websites with human audiences. But how do we get there?

Here we unravel this often confusing and complicated topic to uncover how ad fraud is affecting local markets and what publishers can do to minimize its impact.

Why local markets?

Fraudsters go where the money is. This year in the U.S., mobile ad spending is expected to surpass television advertising. Pair this influx of money with a platform where anti-malware and ad fraud detection is immature or non-existent, and mobile advertising platforms and local markets become prime targets for ad fraud.

How does local ad fraud happen?

Bad actors exploit one of the strengths of local markets: local audiences. Recent research from BIA/Kelsey estimates that out of $76 billion in local ad spending, almost $3 billion is lost to geotargeting ad fraud annually. Marketers want to reach their audiences on their mobile devices when they are in specific locations. Fraudsters take advantage of the appeal of geotargeted advertising by faking a device’s location. This means ads run on “devices” that declare GPS coordinates, but that data can be easily faked, allowing fraudsters to steal from locally targeted ad budgets.

Dr. Augustine Fou, cybersecurity and ad fraud researcher, in partnership with BIA/Kelsey, conducted an experiment that compared mobile ad performance in programmatic campaigns targeting Houston during a hurricane and arbitrary market with a very targeted audience: Butte-Bozeman, Montana.

The ad campaigns specifying these two geographic areas were successfully executed by the programmatic exchanges, but the results were not logical. For example, 100 percent of the ads were loaded on Android devices and all ad impressions were generated at 5 a.m. Houston time and 4 a.m. Bozeman time. These findings, among others, show that the app usage in these campaigns did not follow the patterns of typical human behavior and were the result of fraudulent activity.

This initial test was expanded to include the geotargeting of 16 DMAs through programmatic media buying. While there were variations in the levels of fraud between DMAs, the results showed similar rates of ad fraud (shown in dark red on the following chart) throughout all 16 markets.

“Local publishers have real audiences—the people who live and work in the communities and consume the content they create,” explained Fou. “Marketers need to understand the risks of ad fraud in geotargeted campaigns and consider spending their ad budgets directly with local publishers to avoid those risks.”

What can publishers do?

Legitimate local publishers work hard to create quality content that attracts real human audiences. Publishers who can prove that they have quality, human audiences will be the winners as marketers look to reward those who proactively combat fraud.

Knowledge is power. As more money is spent to reach local audiences, publishers must pay attention to how ad fraud is occurring in their markets and demonstrate that their sites are safe places for advertisers. Understanding how ad fraud occurs and explaining to marketers the issues that make them vulnerable is an important first step, but more action must be taken.

Here are a few steps publishers can take to bring us closer to minimizing ad fraud in the buying/selling process:

1. Follow Best Practices

Know your business practices for monetizing your site, sourcing traffic and choosing vendors, and educate your staff on best practices. Steps to providing transparency include implementing Ads.txt and having a third-party audit.

2. Use Accredited Vendors

Implement tech to prohibit ads from being served to known bots.

3. Actively Manage Media Monetization

Dedicate staff to review traffic patterns and sources.

4. Understand Traffic Sources

Revisit and scrutinize current traffic sourcing practices.

If you want to learn more about what you can do to help, download the AAM white paper Three Truths to Confront the Digital Ad Fraud Crisis. By looking deeper into ad fraud and how transactions take place within the ecosystem, it’s clear that ad fraud – on both national and local levels – is conquerable.