Google Rewrites Spam Rules on May 15 to Explicitly Ban AI Search Manipulation — GEO Industry Now Has a Target on Its Back

What Just Changed

On May 15, 2026, Google quietly rewrote the introductory language of its Search spam policy. The new text, first spotted by Search Engine Land, now reads: "spam refers to techniques used to deceive users or manipulate our Search systems into featuring content prominently, such as attempting to manipulate Search systems into ranking content highly or attempting to manipulate generative AI responses in Google Search."

The final clause is new. Previous spam rules were built around traditional ranking manipulation — keyword stuffing, link schemes, cloaking. None of that language explicitly covered AI Overviews or AI Mode, Google's newer search products that synthesize answers from across the web. That gap is now closed.

What the Policy Actually Covers

According to Gizmodo's reporting, the updated policy enumerates several specific spam techniques now officially targeted:

• Cloaking — showing crawlers different content than what users see
• Expired domain abuse — buying trusted old domains and loading them with low-value content to borrow authority
• Hidden text and links — burying instructions or signals designed purely to game systems
• AI-generated content farms — using generative AI tools to mass-produce pages that add no value for users

Google isn't banning AI-written content outright. It's banning AI content produced at scale specifically to manipulate search systems. There's a distinction Google is betting its enforcement teams can detect.

The GEO Industry Is Directly in the Blast Radius

An entire consulting sector has emerged around "GEO" — generative engine optimization — promising clients they can get their brands regularly cited and recommended by AI search tools.

The tactics are exactly what Google's new policy targets. Biased "best-of" listicles designed to appear authoritative to language models. "Recommendation poisoning" — embedding instructions in web content to cue retrieval systems into treating a site as a go-to source. Essentially, prompt injection disguised as editorial content.

The Verge reported that a BBC journalist demonstrated how easy this was earlier this year, using these exact techniques to get himself ranked as the "best hot dog eating tech journalist" in Google AI search results. The demonstration highlighted a serious vulnerability in the system.

How Enforcement Works

Gizmodo reports Google uses a combination of automated systems and human reviewers to detect violations. Sites found in breach face two outcomes: rank demotion or outright removal from Search results.

Neither is good for a business that depends on organic traffic. Removal is catastrophic.

The policy does NOT define a specific threshold for enforcement — meaning Google retains full discretion on what triggers a penalty. That vagueness is intentional. It's the same playbook Google used with traditional spam rules: broad enough to catch bad actors, flexible enough to adapt as tactics evolve.

What Mainstream Coverage Is Missing

Most coverage of this update frames it as Google "protecting users." That framing is incomplete.

This is also Google protecting its core business model. AI Overviews already cut into click-through rates for publishers. If GEO tactics poison the AI answer layer — flooding it with manipulated citations and fake authority signals — the product becomes unreliable. Users leave. Ad revenue drops. Google cannot afford for AI Mode to become a cesspool of brand-injected recommendations.

This policy update reflects both Google's financial interests in the AI search era and its user experience objectives. Additionally, the enforcement gap between policy and practice deserves scrutiny. Google has published spam policies for decades. Enforcement has always been uneven. The SEO industry found workarounds for every previous update within months. GEO practitioners may do the same — especially since the manipulation techniques here are more sophisticated than keyword stuffing ever was.

What This Means for Regular People

If you run a website or publish content, the rules just changed. Strategies specifically designed to get your brand cited in AI Overviews — not just ranked in blue links — are now officially spam. Act accordingly.

If you're a user relying on Google's AI search answers, this policy is meant to make those answers more trustworthy. Whether enforcement delivers on that promise remains to be seen.

Google drew the line. Now comes the hard part: holding it.