is The Verge’s senior AI reporter. An AI beat reporter for more than five years, her work has also appeared in CNBC, MIT Technology Review, Wired UK, and other outlets.
Google is stepping things up in the AI agent browser wars.
The company is launching a suite of new features deeply embedding Gemini into Chrome. That starts with the announcement that Gemini in Chrome will no longer require a membership fee and will begin rolling out to both Mac and Windows users in the US starting today. It’s all part of the battle for consumer use of AI-fueled browsers, which OpenAI, Anthropic, Google, Perplexity, and other companies are all fighting to win.
Like OpenAI’s ChatGPT Agent, Google is also planning to introduce an ability for Gemini in Chrome to be able to do “tedious tasks” on your behalf in the coming months, Charmaine D’Silva, Chrome’s director of product management, said in a briefing with reporters on Wednesday. It’s designed to grocery shop for you from a grocery list in your email, reschedule deliveries, set up hair appointments, book restaurant reservations, and more, and it will have checkpoints in place for anything that’s considered “high-risk” or “irreversible,” she said. Google did not provide a specific launch date when asked by The Verge.
Other Gemini features are shipping sooner. The new features Google announced Thursday also include giving Gemini in Chrome access to Google Workspace for both regular and Enterprise users, with the rollout starting today, as well as enabling integrations with other Google products, like Calendar, YouTube, Maps, and more. The changes allow Gemini in Chrome to “find relevant information on your screen and also take action on your screen” with those tools, D’Silva said.
“Enterprises are a pretty important focus for Chrome generally,” D’Silva said.
On desktop Chrome, users will now be able to use the Gemini AI agent across multiple different tabs to compare products, summarize information from multiple sources, and recall previous pages from a user’s browser history. That means that starting today, users will be able to can close their myriad tabs and then have the AI agent recall them.
“Say you were looking at team-building activities and it’s the end of your day — if you wanted to pick it back [up] the next day, typically what people would do was leave those tabs open,” D’Silva said. “But now you can close those tabs, and the next morning you can go and say,’ Hey, can you show me those team-building activities that I was looking at yesterday? And we automatically show you.”
On mobile, Gemini was already integrated into Android, but users will now be able to share the entire context of a page, not just what’s currently on their screen, so they can ask “deeper questions,” D’Silva said. iPhone users will be able to access Gemini via the Chrome app soon.
Improvements to how AI agents work in users’ browsers have been coming for a while. Last year, Anthropic introduced Computer Use, allowing Claude as an AI agent to use your browser and complete tasks on your behalf, and months later, OpenAI announced Operator, which was designed to do the same. This past July, OpenAI combined its Deep research and Operator features into one agentic tool, ChatGPT Agent. That same month, Perplexity launched Comet, its own AI-powered web browser. Atlassian just spent $610 million to buy The Browser Company, makers of the AI-infused browser Dia.
