Google Cloud introduces Gemini agent to change enterprise work
SiliconANGLE Kyt Dotson
Google Cloud launched Gemini agent for work across web, mobile, desktop and apps. It can act on its own and even stop when a budget cap is hit.
Based on reporting by SiliconANGLE, Kyt Dotson — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Google Cloud has introduced Gemini agent, a new AI assistant built to work across web, mobile and desktop without a dedicated interface. It can also show up in the places people already use — the command line, Google Workspace, Microsoft 365, Slack and third-party apps — with Google calling it “omnipresent.”
The pitch is broader than a chatbot. Google Cloud chief executive Thomas Kurian said users should give Gemini objectives rather than step-by-step instructions, then come back to finished work. He said the agent can answer questions, handle knowledge work, create images and media, and turn ideas into completed products, all while staying tied to a company’s workflows, systems of record and enterprise controls.
A big part of the design is persistent execution. Gemini keeps one set of memories, context and personalization across channels, and it can spin up smaller worker agents with their own identities, expertise, storage and even @agents.company.com email addresses. Google says those sub-agents only see the context that users or team members provide. The company is also leaning hard into model choice: simple tasks can run on Gemini Flash, longer jobs can use a frontier model such as Argon, and Claude models from Anthropic are available now, with other private and open models coming later.
Google is trying to make the system useful inside real businesses, not just impressive in a demo. Gemini connects to Confluence, Microsoft Office, Teams, Slack, Workspace, Git, Jira, Salesforce, ServiceNow, BigQuery, Databricks, Postgres, Snowflake and even files on desktops. It also ships with a global library of skills, while teams can create and publish their own reusable skills in a shared company registry.
Cost controls sit right alongside the automation pitch. Google said token prices are down about 98% since 2024, even as enterprise AI usage has grown. It is now rolling out flexible spending options, smart routing that picks the right model for each job, and hard budget caps in Cloud Billing Console that can pause Gemini when spend is hit. If companies want the bill attached to a department instead of a giant mystery bucket, Google says that’s built in too.
My take — AI-written commentary, not fact-checked reporting
This is the real enterprise AI sell: not magic, control. Memory, routing, budgets and audit trails matter more than another glossy demo that forgets everything by lunch. The industry has finally noticed that companies will pay for agents that behave like staff, not like chaos with a logo.
Read more about this at: SiliconANGLE
Related stories
Google Cloud’s Always-On Memory Agent Replaces RAG and Embeddings With Continuous LLM Consolidation on Gemini 3.1 Flash-Lite
MarkTechPost · 2 months ago ·
24