Google began rolling out agentic AI features to its Gemini platform on Thursday, giving business customers the first access to software agents that can plan and execute multi-step tasks with minimal human input, the company announced at its annual developer event.
The move puts Google in direct competition with OpenAI, Anthropic, and Microsoft, all of which have raced to ship so-called agentic tools over the past 18 months. Unlike a standard chatbot that answers one question at a time, an agent can book a meeting, pull data from three internal systems, and draft a follow-up email in a single chain of actions.
Google is starting with enterprises rather than consumers, a sequencing decision that reflects where the money currently is. Corporate customers already pay for Gemini through Google Workspace and Google Cloud, giving the company an installed base to sell agents into without acquiring new users first.
CEO Sundar Pichai said at the event that Google still has a strong shot at reaching scale with its agentic efforts, pointing to the company's existing cloud and productivity footprint as the foundation. Pichai has repeatedly framed AI agents as the next phase of the company's core products rather than a standalone offering.
The stakes are high for Google, which has spent the past two years defending its search advertising business against AI-native competitors. Agents that handle workplace tasks could deepen corporate reliance on Google's ecosystem, but they also raise reliability questions — an agent that misreads an instruction can take a wrong action across multiple systems before anyone notices.
Google has not disclosed pricing for the agentic features or a full list of which Workspace and Cloud tiers will receive them first. The company said access will expand to more business customers in the coming months.
TechCrunch Disrupt, where Google's announcement was discussed, opens its doors on Oct. 13, with the final day to book an exhibit table set for Oct. 2. The conference expects more than 10,000 attendees.