Google has been quietly testing a new AI ‘personal agent’ for its Gemini platform. According to a report from Business Insider, Google is internally developing a personal agent called Remi, which will take over OpenGL, a tool on which Sam Altman has spent millions. The tool is reportedly designed to go beyond answering questions or creating content to acting as a 24/7 assistant that can take actions on behalf of users at work, school, and daily life. The internal description seen by BI reads, “Remy is your 24/7 personal agent for work, school, and daily life, powered by Gemini. This turns the Gemini app into a true assistant that can take action on your behalf – not just answer questions or generate content.”Remi is also said to be deeply integrated into Google services, capable of monitoring critical tasks, proactively handling complex workflows, and learning user preferences over time. Employees are currently testing the agent in the employee-only version of the Gemini app.

AI agents will become a major focus

AI agents have become a major focus for leading laboratories as models become more reliable in powering autonomous devices. Google has already introduced “Agent Mode” and related features in Gemini, but Remi appears to be a more advanced step. This latest project from Google also bears similarities to OpenClaw, an independent AI agent that went viral earlier this year for its ability to respond to messages and conduct research. OpenAI CEO Sam Altman later hired the creator of OpenClaw, underscoring the competitive race to create the most capable agent.Remy could position Google as a direct challenger to OpenAI’s agent ambitions. While OpenAI has invested heavily in agent technology, Google’s integration into its ecosystem — from Gmail and Docs to Calendar and Drive — could give Remi a unique edge in everyday productivity.Remi is currently described as a “dogfooding” project, meaning it is being tested internally before public release. Google hasn’t disclosed a launch timeline, but the company is expected to showcase its next wave of AI products at its upcoming I/O event later this month, with agents likely to be the centerpiece.

Google announces I/O 2026 date

Google announced its annual developer conference, Google I/O 2026, will take place May 19-29 at the Shoreline Amphitheater in Mountain View, California. The two-day event will be available both in-person and online and will be live-streamed online on io.google. Google CEO Sundar Pichai also announced the date of Google I/O 2026 in a social media post on X. Pichai wrote, “See you all at Google I/O starting May 19th! http://io.google.”“The event will include keynote speeches from Google leaders, fireside chats, product demos, and more. While you wait for the big day, check out this year’s save-the-date experience, where you can create, create, and remix your way through a playground of experiences built with Gemini,” Google said in a blog post.

Google I/O 2026 will focus on AI breakthroughs

According to the blog post, Google said this year’s showcase will focus on ‘AI breakthroughs’ in Gemini, Android and more.If last year’s I/O is any indication, AI will dominate the agenda. In 2025, Google unveiled Gemini 2.5 Pro and new generative models like Flash, Imagen 4 and VO3, and AI-powered features in Search, Gmail, Chrome and Shopping. Even hardware demos — from an Android XR headset to Gemini-powered smart glasses — were designed around AI.