This is the year that AI went from being a future concern for publishers to becoming an integral part of their daily reality. By 2025, generative AI has been integrated into traffic analytics, licensing negotiations, and product development.
Search clicks have decreased. AI responses are increasingly replacing the blue links that direct Internet users to publishers’ sites. Some companies have entered into AI content licensing deals, while others have blocked AI crawlers, filed lawsuits, or shipped their own AI-based products.
Here are some of the moments that defined how publishers adapted to the AI era this year.
Google AI Insights Reduce Search Traffic
Google’s AI-generated search summaries, called AI Overviews, celebrated their first anniversary in May. Around this time, publishers began quantifying the impact of AI previews on their search clicks, with some reporting 50-90% lower CTRs when an AI summary appeared. In August, Digital Content Next found that AI insights were linked to a 25% drop in referral traffic. The future of “clickless search” began to materialize, and it wasn’t pretty.
During this year’s third-quarter earnings calls, publishing executives assuaged investors’ concerns about the impact of traffic on their businesses and shared their strategies for guarding against traffic erosion by investing in video, direct-to-audience approaches and AI licensing plays.
Editors draw lines on AI crawlers
In the first half of this year, publishers were playing an endless game of Whack-a-Mole with AI crawlers as they tried to stop those who scraped their sites for content used to train their models – without compensation.
But on July 1, Cloudflare launched an AI bot blocking tool that allows publishers and other website creators to block all AI crawlers, as well as have the option to implement a bot payment feature to help publishers monetize AI bot traffic.
Cloudflare’s tool hasn’t ended the AI bot war. But many publishing executives saw it as a way to draw a line in the sand and push tech companies to stop taking down their sites with reckless abandon.
In September, Cloudflare launched a content signals policy for its robots.txt file, which gives publishers a way to communicate how they want and don’t want AI crawlers like Google’s to use their content once it’s removed. Although Google technically separates its search crawler (Googlebot) and its AI crawler (Google-Extended), even if a publisher blocks Google-Extended, their content may still appear in AI previews, as it appears in Google Search.
While this doesn’t guarantee Google’s (or any app’s) compliance, it gives publishers a way to communicate that AI crawlers can scrape their sites to index search pages, but can’t use their content to train AI systems. Publishers like The Atlantic are working on adding this update to their robots.txt.
More tech companies are participating in AI content licensing deals
AI content licensing deals have continued to expand this year, as big tech companies like Meta, Microsoft and Amazon have become bigger players in the market, providing publishers with more opportunities to get paid for their content.
Companies like OpenAI and Perplexity continued to sign deals with more publishers this year, including USA Today Co., The Washington Post and the Guardian.
On Dec. 5, Meta signed seven multi-year AI content licensing agreements with publishers, including CNN, Fox News, People Inc. and USA Today Co., to incorporate their content into its large language model (LLM), Llama.
Two months prior, Microsoft entered the game with the launch of its pay-as-you-go AI content marketplace, signing People Inc. and USA Today Co.
Amazon signed deals this summer with Condé Nast and Hearst for its AI sales assistant Rufus, as well as an AI training deal with The New York Times.
Google has been quieter. In January, Google struck its first AI content licensing deal with a news publisher for its Gemini chatbot with The Associated Press.
Meanwhile, the Really Simple Licensing (RSL) Collective – an AI licensing framework designed to standardize how publishers tell AI systems what content they can use and how they should pay for it – launched in September and now has more than 50 publishers joining its efforts, including People Inc., Ziff Davis, Yahoo, Arena Group, BuzzFeed, USA Today Co and Vox Media.
Prosecutions intensify
Again, more and more lawsuits have been filed against major tech companies for their use of publisher content to train their AI systems.
In December, the New York Times and the Chicago Tribune sued Perplexity for copyright infringement.
Penske Media Corporation sued Google in September over its AI summaries, the first time Google has been challenged in court by a major U.S. publisher over AI search.
And in November, 14 newspaper and magazine publishers won a first round of judgment against Canadian AI startup Cohere for copyright infringement. A judge denied Cohere’s motion to dismiss the lawsuit, and said summaries of news articles could infringe copyright, allowing the claims to move forward.
AI references are increasing – but still insignificant
Traffic from ChatGPT, Perplexity and other AI tools to publisher sites increased this year, although by tiny amounts that did little to offset the decline in search traffic.
This summer, more data was released showing growth in AI SEO traffic from platforms like ChatGPT. Yet for many publishers, this was still only a single-digit percentage of overall traffic, if even that. It became clear this year that publishers could not rely on AI platforms to send the same volume of traffic to their sites as search platforms have historically done so far.
Publishers launch their own big AI projects
As the media industry reeled from what seemed like the sudden onset of a barrage of challenges and opportunities brought by AI, publishers launched their own big AI projects. Large companies were expanding product, technical, editorial and other teams.
For example, in December, the Washington Post launched an AI-powered audio product that can be personalized to listener preferences, and Yahoo introduced personalized AI-powered audio news summaries to its app.
Time launched an AI agent in November, allowing readers to ask questions and interact with Time content, in a number of different languages.
The Financial Times made its “Ask FT” chatbot available to all subscribers in April, allowing them to search the publication’s archives in natural language.
Disappointing Google Search Trial Results for Publishers
Unfortunately for publishers, Google’s search engine remains deeply tied to its AI products.
After months of waiting, the Justice Department’s solutions for Google, revealed on September 2, reached a verdict that seemed disappointing to publishing executives who hoped the solutions would separate Google’s search engine crawler from its AI experiences, such as AI previews and AI mode. Or, at least, force them to provide more data on how these products impact publisher clicks from search.
This would have given publishers more control over where their content appears and what it is used for.
