Overview

• Google is rolling out Gemini inside Chrome to Mac and Windows users in the U.S. (English) without AI Pro/Ultra subscriptions, requiring a signed‑in Google account, with Android support already present, iOS updates coming soon, and Workspace availability in the coming weeks.
• Gemini can read and summarize pages, answer questions about what’s on screen, synthesize information across multiple tabs, integrate with Calendar, Drive, Maps and YouTube, and soon help you rediscover previously visited sites via natural‑language prompts.
• AI Mode is coming to the omnibox later this month in the U.S., enabling optional chatbot‑style searches, follow‑up prompts, and page‑specific suggestions that open in a side panel.
• Agentic capabilities slated for the coming months will let Gemini act on web pages by clicking, scrolling and typing to complete tasks like booking a haircut or ordering groceries, with clear indicators, user takeover at any time, and required confirmation for high‑risk steps.
• Security additions include on‑device Gemini Nano to block fake virus and giveaway scams, reduced intrusive permission prompts, and a password agent that can change compromised credentials with one click on supported sites, with the Gemini button removable from the toolbar if desired.