New Google Test is Rewriting 'Discover' Feed News Headlines with AI

Google is testing a controversial change in how it presents news content within its popular Discover feed. The feed, which provides personalized news and updates across millions of Android phones, is currently running a UI experiment that replaces the original Google Discover headlines crafted by publishers with short, AI-generated summaries.

Google confirmed the ongoing test to The Verge. The company claims that the intent is innocent: to make “topic details easier to digest before they explore links from across the web.” However, the results of this artificial intelligence rewrite are frequently misleading, factually incorrect, or nonsensical. This raises immediate alarm bells across the publishing industry.

Google Discover AI rewritten headlines: The risk of misleading news titles

Journalists spend significant time crafting a headline that accurately summarizes a story while inviting a click. But now, Google’s AI is stepping in and reducing that work to a handful of poorly chosen words. For instance, an article accurately reporting that Valve’s upcoming Steam Machine would likely not be priced like a traditional console was summarized by Google’s AI as “Steam Machine price revealed.” Of course, this claim was clearly untrue.

There was another case of a piece discussing how players in the game Baldur’s Gate 3 found a loophole related to in-game child characters. Google’s AI ruthlessly condensed it to a disturbing and misleading phrase: “BG3 players exploit children.”

This behavior creates a major conflict. News outlets work hard to follow Google’s own quality guidelines. The company’s algorithm often penalizes sites for using deceptive or sensational headlines. When Google’s AI attaches actively misleading clickbait to a publication’s work—whose name appears right next to the false headline—it damages the outlet’s editorial reputation.

Plus, the fact that Google often hides the disclosure that the headline was “Generated with AI” behind an extra click only makes the problem worse.

A larger trend

This latest experiment fits into a growing trend. Google is clearly prioritizing its own AI summaries and interfaces over driving traffic to the open web. Features like the AI ​​Overview in Search already answer complex queries directly on the search results page. This often eliminates the need for a user to click through to the original source. Currently the firm is also testing the merge of AI Mode and AI Overviews in Search.

For publishers, who rely on traffic to fund their operations, this AI-powered headline rewriting test feels like another step toward diminishing their editorial control and marketing agency. A headline is an editor’s primary tool for attracting readers. So, replacing it with an automated, flawed summary is akin to altering an artist’s work.

For now, Google insists this is just a small UI experiment. The question remains whether public and industry backlash will be enough to convince Google that maximizing click-throughs through flawed AI headlines is not worth the cost. Things like journalistic credibility and the wider web ecosystem should continue to be the most important.

The post New Google Test is Rewriting 'Discover' Feed News Headlines with AI appeared first on Android Headlines.