Google's AI is Rewriting Your Headlines. Publishers Are Furious.

Google Just Became an Invisible Editor-in-Chief for the Entire Web

Imagine you spend hours crafting the perfect headline. It’s accurate, it’s punchy, it reflects your brand’s voice. Now imagine Google just throws it away and writes its own.

That’s not a hypothetical. It’s happening right now.

Google has expanded a controversial experiment and is now using AI to rewrite the headlines of news articles directly in its main search results. This isn’t just in the Discover feed anymore; it’s happening in the classic “10 blue links” that have been the foundation of the web for decades.

Publishers are, to put it mildly, furious. And they should be. This is a massive deal.

What Happened

For months, Google has been testing AI-generated headlines in its Discover feed, often with messy results. Now, the company confirmed to The Verge that it’s a “small” and “narrow” experiment in its core search results, impacting news sites and others. They’ve stopped calling it an “experiment” in Discover and now refer to it as a “feature” that “performs well for user satisfaction.”

The AI’s goal, according to Google, is to make titles better match a user’s specific query. But the execution is causing chaos, with the AI often completely changing the meaning and context of the original work.

Here are some of the most glaring examples:

• Original Headline: “I used the ‘cheat on everything’ AI tool and it didn’t help me cheat on anything”

• Google’s AI Headline: “‘Cheat on everything’ AI tool.”

  • The Damage: This turns a critical review into what looks like an endorsement.

• Original Headline: “Valve’s Steam Machine looks like a console, but don’t expect it to be priced like one”

• Google’s AI Headline: “Steam Machine price revealed.”

  • The Damage: This is factually incorrect. The price had not been revealed.

• Original Headline: An article about players finding a loophole with in-game child NPCs in Baldur’s Gate 3.

• Google’s AI Headline: “BG3 players exploit children.”

  • The Damage: A deeply disturbing and misleading summary that strips all gaming context.

Sean Hollister at The Verge put it perfectly: “This is like a bookstore ripping the covers off the books it puts on display and changing their titles.”

Why This Matters

This isn’t just about a few clumsy word changes. This is a fundamental shift in the relationship between creators and the world’s largest discovery platform.

The real story here is control. For publishers, the headline is everything. It’s their brand voice, their editorial judgment, and their primary tool for attracting readers. When Google rewrites it, they lose control over their own work at the most critical point of discovery.

This escalates the already-raging war between Google and publishers over AI. With AI Overviews (formerly SGE) already answering queries directly and reducing click-through traffic by as much as 42% for some publishers, this feels like another nail in the coffin. Publishers are seeing their content used to train and power Google’s AI, only to have that same AI then compete with them and rewrite their work.

For you, the reader, it erodes trust. You can no longer be sure that the headline you see in a Google search result is what the author actually wrote. It inserts a layer of AI-driven interpretation—optimized for Google’s metrics, not journalistic accuracy—between you and the source.

Under the Hood: How It Works

This isn’t a simple title tag truncation. Google’s systems are programmatically analyzing the content of an article and the user’s search query to generate a new title from scratch.

Google’s official line is that for topic-based queries, the AI-generated headline isn’t a rewrite of a single article, but rather a synthesis of information from a range of sites.

Think of it like this:

1. You search: “latest on Steam Machine price”
2. Google’s AI sees: Multiple articles discussing the lack of a price, the design, and speculation.
3. The AI synthesizes: It incorrectly concludes that pricing information is the key topic and generates a definitive-sounding but false headline like “Steam Machine price revealed” to match the perceived user intent.

There is currently no opt-out for publishers. If they want to block Google’s AI from using their content, they have to use tools that would also prevent them from appearing in regular search results—an impossible choice for any business that relies on web traffic.

Here’s a visual breakdown of what’s happening:

graph TD
    A[Publisher writes article with headline 'X'] --> B{Google indexes the article};
    C[User searches for 'Y'] --> D{Google's AI analyzes article content AND user query 'Y'};
    D --> E[AI generates a new headline 'Z' it thinks best matches 'Y'];
    B --> F[Original article is stored];
    E --> G((Show result with headline 'Z' and link to original article));

The most concerning part is the lack of transparency. The AI-generated titles often appear without any clear label, making them look like the publisher’s own work.

What to Do Next

Don’t just get mad. Get proactive.

• For Publishers & SEOs: Your new job is defensive monitoring. Live in Google Search Console. Regularly check how your key articles are appearing in search results for top queries. Document every instance of a rewritten headline with screenshots. This data is ammunition.

• For Developers & Tech Enthusiasts: Be skeptical of every search result. The headline is no longer a guaranteed source of truth. Always click through to the article to see the author’s intended title and framing. Support publications directly through subscriptions or by visiting their sites.

• For Everyone: This is a conversation we need to have. Share egregious examples of rewritten headlines you find on social media. The more light we shine on this, the more pressure Google will face to provide transparency and, crucially, a way for creators to control their own work.