Google's AI Rewrites Page Titles at Scale, Challenging SEO Headline Control

Google has long reserved the right to modify displayed titles, but what's new in 2026 is the scale and intelligence of these rewrites. Rather than simple truncation or pulling anchor text from links, Google's AI is now drafting alternative headlines that it believes will improve click-through rates and better satisfy user intent.

Google confirmed it is testing AI-generated titles in traditional Search results, not just Discover. The test is described internally as "small" and "narrow," and not approved for broader rollout. It impacts news sites but isn't limited to them, with the stated goal of better matching titles to queries and improving engagement. SEOVendor's tactical briefing, published March 23, frames this as a fundamental shift for any agency managing title optimization at scale.

Google is running a live experiment in which AI generates alternative titles for web pages, including news articles, that appear in traditional Search results in place of the publisher's original headline. This is different from Google's long-standing practice of shortening or adapting titles for display purposes. The AI is creating new phrasing, not just selecting from existing on-page text.

The concrete examples reveal why that distinction matters. Google reduced a headline reading "I used the 'cheat on everything' AI tool, and it didn't help me cheat on anything" to "'Cheat on everything' AI tool," stripping the article's entire conclusion. It changed another, "Microsoft is rebranding Copilot in the most Microsoft way possible," to "Copilot Changes: Marketing Teams at it Again." In both cases, the rewritten version erases the editorial framing the author deliberately chose.

Google announced in January 2026 that AI headlines in Discover are no longer an experiment; they're a permanent feature. The stated reason is that the feature "performs well for user satisfaction." Industry observers are particularly wary because of what happened in Discover, where a similar "small experiment" appeared late last year and within a month Google reclassified it as a permanent feature. That precedent gives the current Search test real weight.

By remaining indexed on Google Search, publishers accept terms that permit this. Google offers no opt-out mechanism for individual policies and no way to flag objections to specific rewrites. It is all or nothing.

Agencies face three immediate concerns when Google rewrites client titles: brand messaging gets diluted (a premium service may be reframed as generic or price-focused); CTR attribution becomes murky (was the improvement from the agency's SEO work or Google's AI intervention?); and compliance and legal language disappears, which for clients in regulated industries like healthcare, finance, or legal can create real liability.

The compounding effect is real: publishers already face AI Overviews that answer queries without requiring a click and reduced referral traffic, and now the headline accompanying their blue link, the last tool they have to attract a click, may be replaced by a machine-generated alternative.

Understanding the triggers gives agencies leverage to minimize unwanted rewrites. Google's AI is most likely to rewrite a title when the title tag is keyword-stuffed or reads unnaturally, doesn't match the page's primary content, runs over roughly 60 characters, uses boilerplate phrases across many pages, or targets a keyword not well-supported by the page body.

Setting up ongoing monitoring to catch title rewrites as they happen is now standard practice, using SERP scrapers or platforms like Semrush, Ahrefs, or Screaming Frog to compare title tags against displayed results. The agencies that treat this as a signal about content quality, rather than a bug to fight, will be better positioned if Google's Search experiment follows the same path as Discover and becomes a permanent feature.