Google is testing the use of AI-generated headlines in Search, replacing the original headlines written by publishers. This has raised concerns among publishers, who say Google is changing their work without warning and sometimes altering its meaning.
The experiment, which started in Google Discover, is now showing up in regular Search results, including the well-known ’10 blue links’ that have been a key part of Google’s main product.
Google says the test is small, but the implications are not
According to The Verge, Google is running a small and narrow test where Search rewrites article titles instead of showing the original headlines from publishers.
Google spokespersons Jennifer Kutz, Mallory De Leon, and Ned Adriance said that the test has not yet approved for a fuller launch, but did not specify how limited the experiment is.
Publishers are worried not only about Google shortening headlines, but also about the company creating entirely new ones.
The Verge found examples where its staff’s headlines appeared in Search in versions they never wrote.
For instance, one article originally titled, “I used the ‘cheat on everything’ AI tool and it didn’t help me cheat on anything,” was changed to “‘Cheat on everything’ AI tool.”
The Verge said this rewrite made it seem like the site was recommending a product it actually did not endorse.
Publishers say AI rewrites can distort tone and intent
This difference is important because headlines are more than just decoration. They reflect editorial decisions about how a story is presented, promoted, and interpreted.
Google’s rewrite system can change the meaning of coverage, not just its style, and compared it to a bookstore ripping the covers off the books it puts on display and changing their titles.
Google changed one of its headlines to ‘Copilot Changes: Marketing Teams at it Again,’ a phrase that was not in the original article and did not fit The Verge’s editorial style. This is not just about appearance.
Newsrooms put effort into writing headlines that are accurate, unique, and legally sound. An AI-generated headline can reduce that work to something shorter, less clear, or more focused on getting clicks.
Google says the goal is better matching and more engagement
Lifehacker shared that the main goal is to identify content on a page that would be a useful and relevant title to a users’ query. The company said it wants to better match titles to users’ queries and encourage engagement with web content.
Google also said the test is not specific to news publications, which suggests it may be trying out title rewrites on other types of websites too.
Still, Google’s explanation has not eased critics’ concerns.
Google has changed how titles appear in Search for years, often by shortening long headlines or picking other text from the page.
What’s different now is the use of generative AI to create new wording that was not in the original article. Google said that even if the feature expands, it would not be using a generative model and would not be creating headlines with gen AI, but did not clarify how this would be done.
A broader shift in how Google mediates news
For publishers, this experiment comes at a tense time.
Google once called AI headline rewrites in Discover an experiment, but later made them a feature, saying they perform well for user satisfaction. Because of this history, publishers are taking the Search test seriously, even though Google says it is still limited.
The main issue is not just about changing headlines, but about control. Search has always connected publishers to readers.
If Google keeps changing how stories are presented before users click, it is not just organizing the web anymore—it is shaping how journalism looks when people first see it.
That is why, even if Google says the experiment is small, it already feels more important than a simple design test.
