Loading article…
Google’s June 2026 spam update launches June 24, adding AI answer manipulation to its rules; see how impressions are counted and what it means for SEO.
Google began rolling out the June 2026 spam update on June 24, expanding its spam policies to cover manipulation of generative AI responses such as AI Overviews and AI Mode [1]. The change signals that tactics like buying citations for AI answers will now be treated like traditional spam, prompting SEOs to watch ranking shifts as the rollout settles.
| At a glance | |
|---|---|
| Update launch | June 24, 2026 |
| Scope expansion | AI Overviews & AI Mode |
| Impression definition | Links shown in AI answers, counted only when expanded [1] |
| Branded search share | 55.9% of downstream traffic after ChatGPT recommendation [4] |
| Desktop CTR trend | Top‑position mobile CTR down 2.2 pts; desktop gains below #3 [3] |
Google updated the leading paragraph of its spam policies to explicitly state that “the Google Search spam policies also apply to generative AI responses in Google Search” [2]. The amendment clarifies that any technique intended to deceive users or manipulate ranking—including attempts to influence AI Overviews—falls under the same rules as older spam tactics. This aligns with the June rollout, which Google announced on its Search Status Dashboard and warned may take several days to complete [1].
John Mueller explained that AI “impressions” now refer to the number of times a link appears in an AI Overview or AI Mode result, but only count when a user expands the hidden link [1]. The Search Console report still lacks click data, meaning a low impression count could simply reflect hidden links rather than a lack of relevance. SEOs should therefore differentiate between visible and expandable links when diagnosing traffic changes.
Independent data from Advanced Web Ranking shows a divergent trend: desktop click‑through rates (CTR) rose in Q1 2026, while mobile top‑position CTR fell about 2.2 percentage points [3]. The desktop gains are concentrated below the third position, suggesting they do not offset the broader mobile softness. Meanwhile, Similarweb’s analysis of U.S. desktop panels across finance, travel, and beauty sectors found that 55.9% of downstream traffic after a ChatGPT recommendation arrives via branded search [4]. This underscores the importance of brand visibility, as AI mentions often translate into direct brand queries rather than clicks on the AI result itself.
Google also reiterated that it does not grant third‑party SEO tools access to internal metrics, warning against claims of “special” ways to influence AI answer rankings [5]. This reinforces the principle that effective SEO—good “GEO”—remains the primary path to visibility, even as AI features evolve.
The extension of spam rules to AI answers marks a clear policy shift: Google is treating AI‑generated SERP features with the same rigor as traditional results. As the update settles, the real test will be whether SEOs can adapt their strategies to maintain visibility without resorting to prohibited manipulation.
Coverage is mostly measured — 103 of 115 reports stay neutral.
Every Monday — the token unlocks, Fed dates & catalysts set to move crypto and markets this week. So you’re never blindsided.
Free · 3-min read · one-click unsubscribe
AI-assisted synthesis by the TrendWatcher Editorial Desk · sourced from 3 outlets · Jun 27, 2026 · How we report
They cite higher potential equity rewards, perceived job security risks from layoffs, and a preference for faster decision‑making and direct impact.
It enforces policies that label attempts to manipulate generative AI search responses as spam.
Planted text on user‑generated pages can be inserted into AI‑generated reports, influencing citations even when the text is minimal.
Detecting manipulated content is difficult because the planted text resembles normal user contributions and is embedded in sources the AI tools naturally retrieve.
The research found that proposed defenses either failed to stop the attack or degraded the quality of results, indicating no clear solution yet.