AI's Role In Modern Publishing Trends

The biggest mistake we’re at risk of making about editorial in the age of AI is treating it like content creation/management instead of knowledge governance. Everyone’s still arguing about whether AI will write stories, replace reporters, or flood us with misinformation…and we’re missing the plot. The existential editorial question isn’t “what words end up on the page?” That’s short-sighted. We have to think long-term where the real existential question is: “What gets encoded into the knowledge graph of culture?” AI shifts the role of editorial from gatekeeping stories to curating the training data of society. • Every editorial choice today is tomorrow’s model weight • Every omission becomes a blind spot • Every frame becomes a default worldview In other words: AI makes editorial permanent. Once an angle, bias, or omission is ingested into large-scale training sets, it calcifies into infrastructure. It stops being “a bad article” and becomes “a baseline assumption for every future system.” This is the highest-stakes shift no one’s talking about. We are out here arguing about “AI vs. journalists” but the real risk is if editorial leaders will accept their new responsibility as stewards of training data, cultural priors, and epistemic integrity…or whether they’ll keep acting like editors of text files while Silicon Valley encodes the canon without them. 👉 My perhaps contrarian take on what almost everyone is getting dangerously wrong: In the age of AI, editorial is becoming less about publishing stories and more about deciding what becomes the scaffolding of collective intelligence.

I watched the State of the Union, but found the State of the Bots more engaging. The State of the Bots is a report published by TollBit CEO Toshit Panigrahi that assesses the volume and impact of web scraping and generative AI on publishers. It illustrates the potentially devastating impact of the combination of extracting information from web sites and reformulating it into easy-to-digest answers for ChatGPT. According to the report: –The volume of web scraping doubled between the third quarter and fourth quarter of 2024. –During that same time, the amount of unauthorized web scraping increased by more than 40%. –Even when blocked, apps like Perplexity used unidentified agents to scrape sites. All that adds up to one devastating conclusion: “AI chat bots on average drive referral traffic at a rate that is 96% lower than traditional Google search.” In plain English that means online publishers can expect to receive 96% fewer clicks on their web sites in a new world where most internet queries are answered using ChatGPT. A friend of mine compared this to the impact of Uber on taxis in New York. Before Uber, taxis were plentiful and cheap. Afterwards they largely disappeared. Uber may be convenient, but its more expensive and it wiped out the value of owning a taxi license. In the case of publishers, it’s not clear how the new system would let them generate revenue to pay writers since the current ecosystem for advertising-supported content depends on traffic and that comes from Google. The alternative is to go behind a paywall. Either way, the result is less, reliable and high-quality information available to the public. Tech companies have generally balked at compensating publishers for content – either to be used in training models or to be accessed as a reference to answer questions. Analyst Matthew Scott Goldstein (msg) calls it “The Greatest Heist on Earth.”  He wrote: “Powerful tech companies, from social media giants to artificial intelligence startups, are quietly siphoning off the value from content creators across all industries.” Goldstein says safeguarding copyright is essential and steps need to be taken now. “Once AI models are trained on stolen content, the damage is irreversible. The longer this goes unchecked, the harder it will be to claw back rights.” It’s possible some of the damage will be mitigated by new search designs Google is unveiling which it says will make it more attractive for people to click through to web sites. What the State of the Bot report makes clear is that that future is coming fast.

Media is constantly in a state of having to reinvent itself, driven by technologies that democratize and revolutionize how people collect information. And in 2025, the abundance of information available to people, and the abundance of people publishing information, is completely unprecedented. It would make Martin Luther's head explode to know that a little over 500 years after the Reformation and the printing press revolutionized the concept of information, and the value of it, that we'd be here -- the potential era of artificial intelligence. A generation ago, the internet stole print advertising budgets, spawning a new generation of digital-first publishers that benefited from social media and search traffic. Now, artificial intelligence–based tools like Google AI Overview, Perplexity, and OpenAI's ChatGPT are siphoning off what remains of publisher traffic, providing much of the information users need within their own ecosystems. Since Google introduced AI Overviews in May 2024, the percentage of zero-click news-related queries has increased from 56 percent to nearly 70 percent, according to a new Similarweb report. To take one example, Business Insider, which recently announced it was laying off 21 percent of its staff, has seen its organic search traffic collapse by 55 percent. Of course, the digital advertising market is also being disrupted. Google, which still accounts for 86 percent of search activity in the U.S., has begun testing ads alongside its A.I. results as a way to monetize these shiny new tools. If the past is prologue, the impacts will be significant. According to eMarketer research, U.S. ad budgets allocated to A.I. search will increase from less than 1 percent this year to about 14 percent in 2029, or more than $25 billion. For comparison, Pew estimates that newspaper advertising revenue fell from nearly $50 billion in 2005 to less than $10 billion in 2022. It’s hardly an exaggeration, as my lunch companion may have recognized deep down, that we’re in for a media brand extinction event. So...what do you do if you're a digital publisher – a company that leaned into the promise of Google Search and Facebook traffic, focusing on trading free for scale, audience size for ad dollars? There are new search engines, like ChatGPT, but a recent study found that although traffic referrals from ChatGPT increased 25x year-over-year between May 2024 and May 2025, they’re most likely going to publications that have partnerships with OpenAI, like Reuters and the New York Post. It's hard not to see how this means all the leverage will sit with Sam Altman — and that's not exactly good news for publisher incentives. As counterintuitive as it might seem, the answer is to go smaller, not larger. I dive into why — and how — in my latest for Puck. https://lnkd.in/ew6sytUK