Gemini agent
Also known as: Google universal agent, Gemini universal agent for work, Gemini work agent
Google announced the Gemini agent at its Gemini at Work 2026 conference on October 8. Google Cloud CEO Thomas Kurian framed it with a single line: 'Work now starts in the prompt window.' Rather than a feature inside a specific app, the Gemini agent is positioned as a single interface for all knowledge work: answering questions, drafting content, creating media, and writing and running code, all from one prompt box.
What separates it from earlier Gemini features is its memory model and skills architecture. The agent maintains four kinds of memory: session memory for the current conversation, semantic memory for facts about the user's role and context, procedural memory for how recurring tasks get done, and episodic memory for what happened in past interactions. Skills are reusable sets of instructions or workflows stored as modular prompts; Google ships a global skills library and lets teams publish custom skills to a shared company registry.
Enterprise controls are a central part of the pitch. Each Gemini agent gets its own identity, a limited access scope, and an activity log. IT sets company rules once and they apply across every agent instance. Admins can set per-project spending caps. Industry-specific versions for financial services and legal are in preview, with government, healthcare, and retail planned next. The agent runs inside Google Workspace apps including Gmail, Drive, Docs, Slides, Sheets, Calendar, and Chat, as well as in Microsoft 365 and Slack.