Friend or foe: Navigating the AI revolution in digital publishing
Published: 05 Mar 2025
Whether you’re an AI-optimist or an AI-sceptic, one thing’s for sure: Generative AI is re-writing the rules of digital publishing.
While AOP’s annual trends report showed that digital publishers are leaning towards optimism regarding the impact that AI will have on digital publishing operations and the ultimate profitability of their businesses, we wanted to dig deeper. Of course, there are two sides for every coin, and where there are challenges, there are also opportunities.
We reached out to the AOP Digital Publishing Awards jury to uncover what digital publishers see as the biggest challenges and opportunities created by the rapid emergence of generative AI technologies.
Audiences are increasingly turning to alternatives to search engines for information discovery – and on top of that, new developments such as Google’s AI overviews are further redefining the way the world searches.
“This diversification means users increasingly find answers without ever visiting the publishers who originally created the content,” commented James Hanslip, CEO at Content Ignite, “resulting in noticeable declines in publisher traffic and audience numbers.”
An article from Brookings predicts that search traffic to websites will fall by a quarter by 2026, whilst referencing that 25% of all the web pages developed between 2013 and 2023 no longer exist. After years of carefully tailored SEO strategies, publishers are now facing a future where search traffic to their sites could be severely impacted.
“Perhaps an even deeper concern for the publishing sector is that Generative AI poses an existential threat to the existing advertising revenue model,” suggested JJ Shaw, Managing Associate at Lewis Silkin. “If ChatGPT has already scanned a publisher’s online articles and can summarise these in an instant (perhaps even for free) in response to user prompts, why would the user need to visit the publisher’s website at all?”
Of course, publishers have long been adapting to constantly evolving algorithms as Google tweak and refine their search experience. While this is a new challenge, publishers have already begun the process of diversifying their audience streams in response to the uncertainty around third-party cookies.
“While the shape of the new normal remains uncertain, it’s reasonable to expect a recalibration over time, with publishers potentially receiving traffic from emerging sources like AI-powered tools,” encouraged Content Ignite’s James Hanslip.
AI is developing rapidly – and leaving legislators in the dust in the process. Indeed most of the challenge publishers are facing regarding traffic reduction is caused by LLMs scraping existing web content for material. When we asked Amanda Barnes, Chief Executive at Faversham House Ltd, what she considered to be the biggest challenge posed by AI, she returned that “it has to be the unauthorised use by bad actors of our content even when it is locked down behind a paywall. There’s a need for clear regulation and transparency on who is using what.”
Many of the existing IP protection laws we have in place don’t adequately respond to the way that generative AI tools operate, or, as Richard Jamieson, Chief Commercial Officer at Press Gazzette, put it, “allowing [AI] to summarise and steal original reporting thereby removes the incentive for publishers to invest in it!”
However, we need a collaborative, cross-national approach to tackle the challenge. “In today’s interconnected world, data and content flow freely across regions, meaning that any country’s laws may fall short in addressing the broader implications of AI usage,” explained Charlie Celino, Head of Strategic Development at News UK. “A global regulatory framework would ensure consistent standards, protect users’ rights worldwide, and prevent loopholes that bad actors could exploit in less-regulated areas.”
Publishers are already taking steps to address these challenges however. As Lewis Silkin’s JJ Shaw referenced, “High-profile legal battles have commenced between publishers and AI companies (such as New York Times v OpenAI) over allegations that AI companies have ‘copied’ large volumes of publisher work without permission and without paying a licence fee.”
Cases such as these are laying the legal foundation for how publisher IP should be treated in the years to come but also remind us that there are opportunities for publishers to work with these technologies. Politico, News Corp, Vox Media, and the Financial Times are just a handful of the companies who have signed a deal with OpenAI allowing the company to use their content to train their LLMs.
However, publishers should be cautious about entering an agreement and consider whether this constitutes a short-term or a long-term win for their brands, as the influx of generative AI-created content into the market is its own challenge for publishers, increasing the noise they must cut through to reach their audiences. “ChatGPT et al, plus AI overviews in search engines, mean publishers must be extra vigilant with what they allow LLMs to access the type of content they produce, and where they allow content to be consumed from,” shared Thomas Lake, Product & Technology Director at Infopro Digital.
One of the biggest challenges posed by generative AI is the risk of inaccurate information being shared and amplified at scale. According to Angie French, AI Advisory / Ethical Intelligence Architect at ALF Consulting, as Gen AI gets better at mimicking content – up to and including the use of celebrities in deep fake videos – publishers face two challenges: “verifying the authenticity of sources they use and establishing the authenticity of their own content in consumers’ minds.”
Broadly speaking, we’re seeing digital publishers adopt guidelines to structure the use of AI within their brands, but it’s vital that the challenges posed by AI’s potential for misinformation are kept front of mind. Mun-Keat Looi, International Features Editor at The BMJ, warned about the importance of “keeping track of where people have used AI so [they] always have a ‘human in the loop’ to factcheck.” Jeremy Bennett, Head of Digital Operations at Bauer, mirrored the sentiment, and went further to caution that “traditional publishing requirements such as fact-checking (looking to back up what AI is telling you in ‘analogue’ sources) could likely become a more time-consuming task.”
AI’s tendency to hallucinate has been well-documented, and while AI-based search tools are becoming better at citing their sources, there’s still a long way to go. This can have knock-on effects for publishers, beyond the potential of citing inaccurate information within fresh journalism. According to Hannah Buitekant, Chief Commercial Digital & Strategy Officer at DMG Media, a “lack of engagement with IP owners leads to unverified results and can undermine the integrity of news organisations by serving up inaccurate information.”
In a worst-case scenario, this has the potential for significant damage to your brand reputation by inaccurately attributing misinformation to a publisher.
Arshiya Nazir, Strategy Director – Addressable and Commerce at Dentsu, shared that “many [customers] struggle to differentiate between real and AI-generated media, which puts the pressure on publishers to ensure authenticity, fact-checking, and ethical AI use.”
However, whilst this is undoubtedly a challenge for publishers, it also creates an opportunity to rebuild trust in the reputation of quality digital publishers. “Factual and verified content has never been more important,” shared Dr Rebecca Whittington, Online Safety Editor at Reach Plc, “and with this new challenge there is also real opportunity for legitimate digital publishers to build a secure audience.”
As AI is prone to the occasional bout of hallucinations, digital publishers need to lean into their human journalists as a core differentiator.
“The value of human fact-checking, boots-on-the-ground reporting, and real-time storytelling lies in the authenticity, emotion, and lived experience that only people can provide,” explained News UK’s Charlie Celino. These qualities bring depth and immediacy to stories in a way that models cannot replicate or must wait to summarise after the fact.”
There are always bad actors within digital spheres, and AI tools empower these players just as much as they do publishers.
“GenAI is making malware-based ad fraud more sophisticated, with AI-generated deceptive ads slipping through security filters and bots mimicking real user behaviour,” explained
Nino Stylianou, Head of Programmatic Yield at Mumsnet. “This makes fraud detection harder, increases the risk of malicious ads, and threatens both revenue and user trust.”
Publishers are facing a dramatic increase in invalid web traffic due to AI-powered crawlers, distorting campaign metrics and undermining trust in the digital advertising ecosystem. Publishers need to rapidly develop new tools in their arsenal to effectively identify and respond to these bad actors and to reassure advertisers that they’re trustworthy partners for their campaigns.
It’s not all bad news though – and AI creates new ways for publishers to partner with advertisers to deliver highly-targeted and effective campaigns as well.
“By using AI to analyse and structure our vast content ecosystem, we can help advertisers build more precise, impactful campaigns that reach the right audiences with greater relevance,” explained DMG Media’s Hannah Buitekant. “One area we’re actively exploring is AI-powered categorization and sentiment analysis, which will allow brands to align with positive, sector-relevant content more seamlessly than ever before.”
Our agency judges also highlighted the value of using AI to deliver enhanced advertising campaigns. Dentu’s Arshiya Nazir shared that “Dentsu has already seen the benefits of AI for this with our Contextual Intelligence offering in placing ads around relevant content, ensuring they are contextually aligned with the content being consumed, thereby improving targeting and engagement.”
Seun Odeneye, MD for Media at KINESSO UK&I, echoed the sentiment: “Today’s most effective advertising is all about personalisation, and GenAI enables publishers to tailor their content for each visitor’s experience. In an era of hyper-personalisation, data is key. Knowing where users are coming from, both upstream and downstream, allows publishers to deliver the most relevant content.”
“One exciting opportunity for digital publishing with the rise of GenAI is the ability to scale content creation like never before, highlighted Chloe Davies, Founder & CEO at It Takes A Village Collective. “AI can help generate drafts, suggest headlines, and even tailor content for different audiences in real time. This means publishers can experiment more, personalise experiences, and respond faster to trends without burning out their teams.”
Where publishers have developed numerous strategies to offer more curated, personalised experiences for their audiences, AI opens the door to new possibilities that can be rolled out across their audiences. As ALF Consulting’s Angie French explained, “publishers can now develop adaptive content that responds to individual reader preferences, learning styles, and knowledge levels - essentially creating "dynamic editions" of the same core material, and crucially at scale.”
Infopro’s Thomas Lake also referenced the ability to serve different flavours of articles based on user preferences, suggesting that “for B2B publishers, this could mean offering ‘lite’ versions of articles for time poor execs needing the key facts, alongside longer in-depth versions for specialists needing to know the detail.” Mumsnet’s Nino Stylianou provided an example from the B2C world, stating that “as a social platform for parents with user-generated content, GenAI can help generate highly relevant, personalised content recommendations, increasing user retention and engagement.”
Beyond this, AI also provides publishers with the opportunity to revolutionise their product development to serve more complex content or to service the needs of a smaller niche of their audience. “The ability to scan, assimilate and analyse complex data points quickly will allow our journalists to contextualise this information, overlaying it with their expert view and deep market knowledge,” enthused Faversham House’s Amanda Barnes. “We’ll be able to produce products we’ve thought about and either hadn’t had the capabilities to monetise or it’s been prohibitively expensive to do so.”
Alastair Lewis, Founder at Quested Consulting, proposed that we consider AI as “assisted intelligence,” a tool that allows us to work smarter and more effectively. However, you have to be strategic about how you approach AI and ensure that your approach is aligned with your organisational north star and brand values.
It is also essential that we work in tandem to develop an industry-wide response. “Publishers must act decisively,” continued Alastair Lewis, “collaborating to define a new normal that not only protects but also strengthens the digital publishing ecosystem.”
Here at AOP, we look forward to supporting publishers as they continue to evolve their approaches to AI, and to seeing more examples of how AI is being integrated into publishing operations in the entries to the AOP Digital Publishing Awards.
The AOP Awards are open for entry! Submit your entry by March 20th for your chance to impress our independent jury of advertisers, agency leaders, and your digital publishing peers.
