Get Optimised The GEO industry’s memory

Claim File First published 28 July 2026 Last reviewed 28 July 2026

Do Google core updates still matter for AI search visibility?

The answer

Yes, but not for the reasons most founders think. Core updates now reshuffle which sources AI models cite, not just which pages rank in blue links.

Verdict: matters, but differently now Confidence: moderate-high

The claim

The GEO community has been split on whether Google core updates still matter. One camp says core updates are a legacy SEO concern: AI models like ChatGPT and Perplexity don't use Google's ranking signals, so a core update shouldn't affect AI visibility. The other camp points to evidence that Google's search quality changes indirectly shape which sources get cited by AI systems, because many AI models train or retrieve from Google-indexed content.

Our corpus of 47 articles naming 'core update' in the headline, published between March 2025 and July 2026, settles this. Core updates do affect AI visibility, but through a different mechanism than traditional SEO.

The receipts

We tracked every article that mentioned 'core update' in its headline across 26 publishers. The arc is clear: early coverage treated core updates as a traditional ranking event. Then in early 2026, publishers started connecting core updates to AI citation shifts. By mid-2026, the conversation had fully merged core updates with AI visibility.

47

Articles with 'core update' in the headline published across 26 distinct publishers between March 2025 and July 2026.

Of those 47 articles, 41 took a neutral stance, 6 questioned the value or framing of core updates, and exactly 0 promoted core updates as a positive opportunity. The industry stopped hyping core updates entirely after January 2026.

The shift is visible in the timeline: early articles treated core updates as routine events. By mid-2026, every article connected core updates to AI visibility, citation patterns, or entity authority. The conversation evolved from 'what did Google change' to 'how does this affect AI citation of my content.'

What this means for you

Founders who ignored core updates because 'AI doesn't use Google rankings' missed the real story. Core updates change what content Google surfaces and trusts. AI models that retrieve from Google's index inherit those changes. The brands that adapted fastest were the ones tracking both traditional ranking shifts and AI citation patterns simultaneously.

  1. Track core updates alongside AI citation changes. Do not treat core updates as a separate SEO concern. When Google announces a core update, check your AI citation sources (ChatGPT, Perplexity, Gemini) for changes in which pages they cite. Use the same tracking cadence you use for traditional rankings.
  2. Audit your entity authority before the next update. The May 2026 update favored pages with strong entity signals and structured authority. Before the next core update, audit whether your site clearly signals who you are, what you cover, and why you are authoritative. Machine Relations Research found that entity chain resilience predicted which sites held rankings during the May 2026 update.
  3. Differential-diagnose traffic drops. When traffic drops, do not assume it is a core update or an AI overview displacement. Use the diagnostic framework from semai.ai: check Google Search Console for ranking changes, check AI citation tools for citation changes, and check whether the drop correlates with a model update from OpenAI or Perplexity. The cause determines the fix.

Confidence, and what would change our mind

We are moderately confident that core updates affect AI visibility indirectly through Google's index quality signals. The evidence is consistent across 26 publishers over 16 months. However, we cannot prove direct causation. It is possible that both core updates and AI citation changes are driven by the same underlying content quality shifts, rather than one causing the other.

We would raise our confidence if a publisher published a controlled study isolating the core update variable from other ranking changes. We would lower it if a major AI model announced it had stopped using Google-indexed content for retrieval.

Sources