Breaking News
Breaking: Google Launches Gemini Agent, a Universal AI Agent for Work
Google Cloud announced the Gemini agent on Oct 8 — one universal AI agent for work with multi-model routing (including Claude), persistent cloud execution, and industry versions.
Google just entered the workplace AI-agent race in a big way. On October 8, 2026, Google Cloud announced the Gemini agent at its Gemini at Work 2026 event — a single, universal AI agent designed to handle knowledge work across the business apps companies already use. Google Cloud CEO Thomas Kurian made the announcement in a keynote and on the company's official channels, with coverage from Reuters and Neowin.
What we know
- One agent for everything: The Gemini agent answers questions, handles tasks, creates content, and writes code from a single prompt box — no separate interfaces per job.
- Persistent execution: It runs in the cloud with a single set of memories, so jobs lasting hours or days keep running after you close your laptop, returning finished work in documents, inboxes, and developer environments.
- Multi-model routing: Each job runs on the best-fit model. Right now that means Gemini models plus Anthropic's Claude models — a notable admission that Google will route work to a rival's models when it fits better, with more models to come.
- Works everywhere: Accessible via web, mobile, desktop, and command line — inside Google Workspace (Gmail, Docs, Sheets, Calendar) and through Microsoft 365 and Slack, not just Google's own apps.
- Connects to enterprise systems: Connectors cover Salesforce, Jira, ServiceNow, BigQuery, Snowflake, desktop files, and any Model Context Protocol (MCP) server. Teams can add reusable "skills" — templates for multistep tasks — shared in a company registry.
- "Coworker agents": Users can spin up sub-agents that act as team members, each with its own email address and access limited to the information they're given.
- Cost controls: Built-in spend caps (set in Cloud Billing) pause work if a cap is hit, with costs tracked per project for department chargebacks.
- Industry versions: Agents built for financial services and legal are now in preview; versions for government, healthcare, and retail are coming soon.
What's unconfirmed
- Pricing and rollout: Google has not published detailed pricing or a full availability timeline in the announcement coverage we've seen — treat "available now" claims from secondary outlets cautiously until the company confirms specifics.
- Real-world performance: The multi-model routing and persistent memory are architectural claims from Google's own announcement. Independent evaluation of how well they work on real cross-system tasks is still to come.
Why it matters
This completes the agent-platform trifecta of late 2026: OpenAI's always-on "dots" agents (September), Meta's personal agent Muse, and now Google's enterprise agent — all racing to be the AI that does your work for you. For employees at companies on Google Cloud, the interesting shift is AI that acts inside the apps you already open, rather than in a separate chat window. And the decision to route jobs through rival models — including Anthropic's Claude — signals that Google sees the agent layer, not the model, as the real battleground.
Sources: Reuters, Unite.AI, MarkTechPost.