As the technology continues to develop at breakneck speed, publishers are taking every imaginable stance on AI. From The Telegraph forbidding staff from using AI-generated text in its journalism, to Politico actively optimizing its website for generative AI crawlers, a wait-and-see approach is not an option.
There are still many unknowns around both the use of generative AI technology and the potential legal and ethical implications. So, it is understandable that many media executives are hesitant to take a stance. But in all likelihood, you’ve got staff using it already; one survey found that 70% of workers did not disclose their ChatGPT usage to their boss.
So how can senior leaders work to introduce AI in a way that benefits the organization, or at least empower the experts and evangelists to explore and share solutions? Here, we look at publishers who have created frameworks for experimenting in ways which benefit the businesses, as well as some tips from an innovation expert to help shape your own AI strategies.
“Waddle Inn” at William Reed
William Reed is a data and events publisher focused on the food and drinks sector. They have found success with their “Waddle Inn” project, named lightheartedly after the way ducks and penguins waddle together, with the extra ‘n’ to make it sound “more like a pub,” explained Chief Digital Officer John Barnes.
William Reed has three locations in the UK, as well as offices in France, Chicago and Singapore, not to mention staff working remotely. So they decided that an online forum would be the best way to bring everyone together. They set up a Microsoft Teams group composed not just of techies and experts, but also those who are new to AI, intrigued, or even frightened by it.
“Whatever your starting position is, here’s a group where you can come together.” Barnes explained. “We use it to share articles, pose questions to each other, we use it to discuss and come up with ideas. It helped form our AI statement… What’s come out of it is a whole series of ideas, some of which we’re working on, some of which we’re still debating, and some of which we’ve just put a spike through because it was too crazy to even contemplate!”
A year on, the Teams group has now become a “record” of William Reed’s AI journey. Barnes noted that some of the ideas they discussed when it first started now seem “less crazy” given the speed at which the AI landscape is developing. “It’s such a useful social resource that seems to be well-liked and well-attended,” he said. “It’s very, very active.”
Immediate Media’s AI experimentation days
Magazine and special interest consumer publisher Immediate Media have found that offering staff events focused on AI has driven experimentation and innovation. They started off with an AI Immersion Day in the summer, which Roxanne Fisher, Immediate’s Director of Digital Content Strategy said was about setting a level playing field of understanding.
“We had a morning of keynote speakers come in and talk about things like the ethical implications and the opportunities, and we also had our CEO stand up and say, ‘This is what we’re doing with AI, and this is what we believe we shouldn’t do,’” she explained. “Then in the afternoon we had really hands-on workshops about prompting, ChatGPT, basic frameworks, Midjourney and how to decide what the right tool is for your use-case.”
But they didn’t just leave it at that. To encourage practical outcomes, Immediate then set up an AI Experimentation Days where they asked staff to submit projects that they might want to use AI for. From over 70 submissions, they took forward 27 projects to work on a hackathon-style event.
“The experimentation days were really useful because people were getting to work on something they had submitted, and were quite excited about,” Fisher said. “But also we had expert facilitators going around the business for the two days. So, if you got stuck…we had people to give really good prompting tips.”
The days were seen as a success by staff as well. Fisher noted that having no expectations around experimentation or productivity gains really helped take the pressure off. Although they’re planning more days in the future, she acknowledged that they’re now working on ways to build the outcomes into the business more strategically. “The experimentation is going really well, but building that into a process or ways to use these tools on a regular basis that’s really meaningful is more difficult,” Fisher said.
Have a clear company-wide generative AI statement
Plenty of publishers have put out statements and policy guidelines to clarify their approach to AI tools over the past year or so. This shouldn’t be seen as a frivolous exercise or a PR move; both Barnes and Fisher noted that publishing clear, public guidelines have really helped reassure staff, as well as provide a framework for internal experimentation.
“If I hear somebody in our company saying, ‘I’m really worried about AI, I’m going to lose my job,’ I don’t see that as a problem. It’s something I need to try and address so that they, for their own well-being, aren’t feeling frightened and worried,” William Reed’s John Barnes said. “We’re not going to be doing certain things, hence having a statement that’s transparent and published that everyone can see.”
Immediate has also put together their own manifesto for how they will use AI, and have communicated that to staff. “The first thing we wanted to do is be really clear in the business what our stance was for using it; what we would and wouldn’t do,” Fisher explained. “We really want to stay up to date and understand it, and understand how it’s going to change everything, but we also want to really maintain the trust of the brands.”
For Fisher, it was important to put guardrails and rules in place for experimentation. This includes basics like not putting personal or sensitive data into tools, but also ensuring staff feel secure. “We did a lot of psychological safety work around it, and tried to make sure people felt they could have the conversations and be really honest about how they’re feeling about it,” she said. “But also lots of reassurance and having that clear stance of what we did and didn’t do was helpful. It set out to people that we’re not about to churn out hundreds of news articles or anything. That’s not what we’re about.”
Lead AI innovation from the top
The success of AI initiatives like these are heavily predicated on support from senior executives. Fisher found that, at Immediate, early encouragement to try and use the technology didn’t take off until more structure was provided. “We realized that the testing and usage needed to come from [the leadership team] more,” she explained. “So we were more prescriptive with people about what we thought they should test.”
Immediate now has an “AI champions board” of people from different disciplines around the business who are passionate about, or interested in the technology. This helps galvanize others. It also means that there are recognized representatives staff can go to with ideas or concerns.
Barnes explained that he sees three types of people who experiment with new technologies in a business: the over enthusiastic early adopters, the pragmatists, and those who are afraid they’ll lose their job to it. “You want to rein in the early adopters who are going to waste lots of time, and you want to make sure that the people that are being pragmatic about what it could or couldn’t do aren’t seen as being the party poopers,” he outlined.
But AI policies and strategies aren’t something that can simply be delegated out to someone else. Media consultant Ian Betteridge, who has worked in leadership positions at publishers like Bauer and Dennis, says that it needs to come right from the CEO. “Firstly, you need to understand this new technology,” he said. “You need to understand how it impacts your ways of working. Because quite often, when it gets to a certain level, everyone just assumes that everybody else knows, or everybody else can just find it all out themselves.”
This means going through several change management processes. “You’ve got to run, as the CEO, a change management process with your executive team,” explained Betteridge. “They then have to run the change management process with the senior leadership team…and they then have to run it with the broader business.”
While that’s going on, media leaders have to “stop people just going off and doing it themselves.” “You’ve got to put guidelines in place from the off,” Betteridge advised.
That doesn’t mean media leaders need to understand the technical detail, or every use case. “But we need to understand it to a point to be able to say to people, that’s not possible, or how to help them unlock things,” Immediate’s Fisher emphasized. “One of the biggest things that people in leadership can do is actually just get hands on with it… therefore you can make sure what you’re asking people in the company to do is realistic and reasonable.”
“Somebody on the leadership team that is looking after digital in some way, shape or form needs to own [it],” Barnes echoed, noting that it can’t simply be delegated to the IT team. “And if they’re not able to lead it on their own, they need to put a group of people around them from their teams or from around the company that can own it.”
Get your AI experimentation going
Some publishers will have teams and individuals eager and ready to experiment with AI. But for those at organizations with leadership teams who are more cautious about the technology, there are still ways to get things moving.
For organizations without strategies in place or a willingness to experiment, “the best way is to always just be able to share results. That’s the thing anybody sitting at the top is going to care about,” Betteridge advised. “So if you can share results and say, we’re now getting five exclusives and more traffic through there, then that matters.”
Barnes suggested simply talking about it as a way to get things started, even if full understanding isn’t there yet. “You want to make sure you’re working within the law, you’re doing something that isn’t going to damage your brand, you’re using tools in the right way to be efficient but not at the expense of humans,” he said, noting that these are points most leaders can agree on.
Whether organizations choose to appoint a board, hire or earmark individual champions, or bring everyone into the conversation, AI is not an issue media leaders can afford to stay silent on. It is a given that there are people within nearly every media organization using generative AI tools in some way. While you might not hesitate for staff to use Grammarly, you’ll likely feel quite differently about using ChatGPT for research or writing. Therefore, it is not only critical that you get started, the impetus and innovation need to start at the top with clear leadership driving smart experimentation.