Google Cloud used its Gemini at Work 2026 event on Oct. 8 to introduce the Gemini agent, which CEO Thomas Kurian called a “single, universal agent for work.” Kurian’s keynote, published as a Google Cloud blog post, also covered a set of announcements on data, industry editions, security and cost.
Google said nearly 500 Google Cloud customers each processed more than 1 trillion tokens in the past year. It says nearly 80% of Google Cloud customers use its AI products and nearly 90% of the Fortune 100 use Gemini Enterprise. Google CEO Sundar Pichai said at the start of the event that Gemini has more than 1 billion monthly active users, TechCrunch’s Sarah Perez reported.
Inside Workspace and data tools
Gemini now works inline in Gmail, Drive, Docs, Slides, Sheets, Chat and Calendar with the same memory and controls it has elsewhere, Google said. In one example from the post, a user asks Gemini to set up a meeting with “the usual team” of regional event leads, and Gemini works out who they are from a chat space and an old thread, then checks calendars to start an email thread. Teams can also create a coworker agent with its own Workspace account, calendar and Drive.
For data work, Google announced skills that turn plain-language questions into saved reports built on BigQuery. Once saved, those reports can be rerun without token costs, according to the post.
Engineers can describe an outcome and have Gemini write PySpark code, train models and fix pipeline problems. Google said Bloomberg Media raised its SQL query accuracy by 63% during early development by grounding its data agents in Google’s Knowledge Catalog, a store of company data definitions.
Industry editions are in preview for financial services and legal, with government, healthcare and retail listed as coming soon.
Google says the financial version has more than 50 built-in skills and draws on data from FactSet, LSEG, S&P Global and SEC filings. It named CME Group and Deutsche Bank as users. The legal version inherits permissions and ethical walls from document systems such as NetDocuments and iManage.
Governance and cost
Google framed agent governance around four questions, starting with who an agent is and what it may do. Each agent gets its own cryptographically attested identity with least-privilege permissions, and every action goes into an audit trail attributed to the agent, Google said. Agents run inside a sandbox, and all of their traffic passes through Agent Gateway, which Google describes as an AI network firewall that enforces company policies in real time.
On cost, Google said per-token prices have dropped 98% since 2024 while enterprise AI volume has grown. Its new controls include smart routing across models and spending caps in the Cloud Billing Console. When a cap is hit, that project’s agent pauses until someone resumes it.
Google also said its latest TPU 8i system delivers 80% better price-performance than the previous generation. It listed its model lineup as Argon for frontier reasoning, Flash for speed, Omni for media and Gemma for open-weight edge work.
The post includes customer claims, all Google’s own figures. It says Bradesco cut some document reviews from an hour to five minutes, Orange Spain has deployed more than 1,000 custom agents and SOMPO built more than 10,000 across its 34,000 employees. It also says an internal agent at the Australian retailer Bunnings saved staff half a million hours of administrative work.
Source: Google Cloud Blog, Google Cloud, “Gemini at Work 2026: Introducing Gemini agent”
Additional sources: TechCrunch, Sarah Perez, “Google brings agentic AI to Gemini, starting with businesses”
