Gemini moves into action

Google announced at a Cloud event on Thursday that Gemini AI can now take ownership of assigned tasks, not just answer questions, all from a single interface.

Gemini's scale in business

Gemini has over one billion monthly active users. Google CEO Sundar Pichai noted that nearly ninety percent of Fortune one hundred businesses now use Gemini Enterprise at work.

Starting with businesses first

Google will initially focus on bringing the agent to businesses before later rolling it out to consumers. This allows the company to solve harder problems around security, scale, and performance.

How the agent works

The agent can be given objectives, not just instructions. It plans work, uses custom skills and tools, and connects to businesses' internal systems like Google Workspace, Microsoft three sixty-five, Slack, and Jira.

“objectives, not just instructions”

Model choice and flexibility

By default the AI picks the best model to complete tasks, but users can choose models including Anthropic's Claude. Google will expand the model picker to include open source and other private models.

The agent as coworker

The AI will have its own Workspace account with an email address and context. It knows team assignments, time zones, approval chains, and calendars. Users can tag it in chats or email it directly.

Early testers and access

Early testers included On, Shopify, and PayPal. The agent will be accessible from iOS and Android, Windows and Mac, the command line, Google Workspace, Microsoft three sixty-five, ServiceNow, and Slack.

Cost control for enterprises

Google introduced flexible spending options including multi-model orchestration, smart routing, and real-time spend caps to help businesses keep enterprise AI costs under better control.