Google Core Update Hits During I/O AI Shift

Welcome to this week’s Pulse, where major shifts across Google Search, rankings, AI Mode, and agentic tools all landed in the same window. From a core update rollout to a major Search interface overhaul and fresh AI usage data, it’s been a packed week for SEO professionals.

Google has officially started rolling out its May 2026 core update, announced via the Search Status Dashboard on 21 May. This marks the second core update of the year and the fourth confirmed ranking update in 2026 so far. The rollout is expected to take up to two weeks, with no detailed guidance yet from Google on what specifically has changed.

The timing is significant, arriving during Google I/O week, where multiple Search-related announcements were also made. This overlap means ranking fluctuations over the coming days may be harder to attribute, as algorithmic updates and product changes intersect. SEO practitioners are being advised to compare performance against pre-update baselines and wait until the rollout fully settles before drawing conclusions.

Industry reaction has already begun. Some SEO professionals have linked the update to broader AI integration within Search, including Gemini-powered features. Others have speculated it may be aimed at sites attempting to optimise specifically for AI citations, though Google has not confirmed any such intent.

Alongside the core update, Google used I/O to reveal a major redesign of the Search experience. The Search box itself has been rebuilt with AI functionality, allowing longer, more conversational queries and supporting multimodal inputs such as images and files. Google described it as the most significant upgrade to the Search box in more than 25 years.

The update also introduced deeper AI Mode integration, with Gemini 3.5 Flash now acting as the default model. Search is increasingly shifting towards AI-generated responses, while new “information agents” are being developed to continuously monitor and summarise the web. Features such as agentic booking and generative interfaces signal a broader move towards AI-led interactions rather than traditional search results.

From an SEO perspective, this shift raises questions about visibility versus clicks. As queries become longer and more conversational, content optimisation based purely on short keywords may become less effective. Instead, content that supports research, comparison, and decision-making may hold more value within AI-driven summaries.

Google also published new usage data on AI Mode, offering its first detailed look at behaviour after a year in operation. According to the data, AI Mode now serves over a billion monthly users, with query volume continuing to grow rapidly. Searches are significantly longer than traditional queries, and follow-up questions are becoming more common. A notable share of searches now include multimodal inputs such as voice or images.

Planning-related queries are also growing faster than overall usage, suggesting users are increasingly relying on AI for research and decision-making before taking action. However, much of this data is based on internal reporting, and independent verification remains limited.

SEO professionals have highlighted a growing challenge around measurement. Without full visibility into AI Mode performance data, it becomes harder to track impact or optimise effectively. This has sparked debate around whether existing analytics tools are keeping pace with changes in search behaviour.

At the same time, Google’s guidance around llms.txt has created confusion. Different teams within Google appear to be offering mixed signals, with Search documentation suggesting it is not required, while Lighthouse includes checks for it. This has led to uncertainty among SEOs about whether such files are necessary for AI-related discovery or agent-based browsing.

Google representatives have attempted to clarify that markdown-style pages may be useful in certain documentation contexts, but are not essential for most websites. However, the lack of alignment across teams has left some ambiguity around best practice, particularly as agentic browsing tools continue to evolve.

Overall, this week highlights a clear direction of travel for Google Search. Core ranking updates continue in parallel with major AI-driven product changes, while the search experience itself is shifting towards conversational, agent-led interactions. For SEO professionals, the focus is increasingly moving from traditional rankings alone to broader visibility within AI-powered systems.