Google update targets spam tactics manipulating AI search responses

Google’s June spam update finished rolling out across all languages on June 26, and the target has widened from classic web spam to tactics that try to manipulate AI responses inside Search. Google’s Search Help says spam covers techniques used to deceive users or manipulate Search systems into ranking content highly, and the company says it can manually remove or demote content after review.

That matters because AI Mode does not live outside Search. Google says it relies on Search’s deep understanding of web information, uses high-quality web content to improve factuality, and can fall back to web links when confidence is not high enough. The company also warns that AI responses can include mistakes, which leaves more room for bad actors to try to shape what gets surfaced before a response is generated.

The sharpest example comes from Cornell Tech’s preprint Deep-Research Agents Can Be Poisoned via User-Generated Content, submitted on May 22, 2026. The paper says deep-research agents repeatedly pull the same user-generated content pages from platforms such as Reddit and Wikipedia. Coverage of the work said roughly 17% to 23% of retrieved web pages in the tests came from user-generated sites, and that adding about 13 words of promotional text to a single source could steer recommendations in a sizable share of runs. That is the problem Google now has to police: one planted comment or recommendation can travel farther than the page it sits on.

For agencies, the line between optimization and abuse is getting narrower inside AI-mediated search. Keyword stuffing and link schemes are still spam, but the bigger risk now is surrounding content ecosystems that can be read as manufactured authority by retrieval systems. If a brand’s content is propped up by fake praise, coordinated commentary, or other synthetic influence, the system may not see a clever growth hack. It may see manipulation.

Google’s scale makes the shift harder to ignore. On June 3, 2026, the company said AI Overviews had more than 2.5 billion monthly active users and AI Mode had surpassed one billion monthly users. That same day, Google introduced a new Search Console control, along with performance insights and updated best practices for website owners navigating AI in Search. Google also said its March 2024 spam update, combined with earlier changes, was expected to reduce low-quality, unoriginal content in search results by 45%, up from an initial 40% estimate.

The agency playbook is blunt: build evidence-backed pages, keep authorship and sourcing clean, and strip out anything that looks like artificial endorsement. In AI Search, fake authority is not a shortcut. It is a liability.