Google is rolling Gemini 3-powered AI directly into Chrome, putting agentic AI inside the browser that 3 billion people open every day. The update adds a persistent Gemini sidebar, Auto Browse task execution, and connections into Gmail, Search, YouTube and Photos through Google’s personal intelligence system.
Chrome holds roughly 63% global browser share. Dropping agentic AI at that distribution is a scale moment the AI safety field hasn’t seen before. Auto Browse can open sites, click links, and fill forms on a user’s behalf based on natural-language goals. The Gemini sidebar reads the content of whatever tab group you’ve pinned.
Regulators are already moving. The EU Digital Services Act treats very-large platforms differently on AI risk, and Chrome’s new capabilities push it further into “gatekeeper” territory under the Digital Markets Act. The FTC’s order against OpenAI over child safety signaled the Commission’s appetite for consumer-AI scrutiny. Browser-level AI that autonomously drives web actions is a new enforcement frontier.
Privacy is the other pressure point. Feeding personal intelligence from Gmail and Photos into Chrome blurs the data-sharing walls that consent frameworks like GDPR and California’s CCPA still assume exist.
Google’s pitch is convenience. The compliance question lands harder at 3 billion users than at 3 million.