Google Chat becomes an agent interface for Workspace

At a glance:

• Google Chat is now the primary UI for interacting with Gemini‑powered AI agents across Workspace.
• The "Ask Gemini in Chat" feature, generally available, integrates with Asana, Jira and Salesforce and offers daily briefings, document generation and meeting scheduling.
• New Workspace Intelligence work graph powers Gemini in Docs, Sheets and Chat, while keeping customer data private unless explicit permission is given.

Google positions Chat as the hub for AI‑driven work

Google Cloud CEO Thomas Kurian announced that the company is turning its Chat app into the central place where employees converse not only with each other but also with AI agents. In a briefing he said, “We’re making Chat the centerpiece through which people talk, not just with other people, but with all their agents, whether they are in Workspace or built on our platform.” The move signals a shift from traditional email and document‑centric workflows to a conversational model where AI can surface information, draft content and even act on a user’s behalf.

The rollout is anchored by the Ask Gemini in Chat feature, which is now generally available. This capability lets users ask Gemini to pull together a “daily briefing” that highlights action items, unread message threads and pending tasks. Beyond summarising, Gemini can generate Google Docs, Slides, schedule meetings, retrieve files from Drive and even execute commands in third‑party tools.

Third‑party integrations broaden the agent’s reach

Ask Gemini in Chat already supports a handful of popular productivity platforms, allowing the AI to act inside:

• Asana
• Jira
• Salesforce

These integrations let the agent create tickets, update project status or log sales activities without the user leaving the chat window. According to Google’s blog post, more connectors are on the roadmap, aiming to cover the full spectrum of enterprise software that workers rely on daily.

Gemini upgrades spill over to Docs and Sheets

Google unveiled additional Gemini‑powered enhancements for other Workspace apps at Google Cloud Next. In Sheets, users can now build and edit spreadsheets using natural‑language prompts. Gemini can orchestrate “multi‑step construction,” pulling data from a user’s files, emails, chats and the web to populate tables, run calculations and generate charts.

In Docs, Gemini gains the ability to create data‑driven infographics, automatically selecting visual styles that match the underlying business metrics. The goal is to let knowledge workers turn raw numbers into presentation‑ready graphics with a single request.

Workspace Intelligence: the work graph behind the magic

Both the new Chat agent and the Gemini upgrades in Docs and Sheets rely on a newly introduced work graph called Workspace Intelligence. This graph connects data stored across Gmail, Chat, Drive, Calendar and even external web sources, learning the semantic relationships between a user’s communications, files and activities. Google says the system “understands complex semantic relationships” and personalises outputs by remembering past work patterns.

Importantly, Google stresses that Workspace Intelligence data is not reviewed by staff, sold to advertisers, or used to train its models unless a customer explicitly opts in. This privacy stance is meant to allay enterprise concerns about data leakage when AI agents have deep access to internal information.

Analyst perspective: a direct challenge to Microsoft Teams

Irwin Lazar, principal analyst at Metrigy, views the move as Google’s bid to compete more directly with Microsoft Teams. He notes that chat has become the primary interface for employee engagement, with email usage declining. “Ideally, I’d like to see the AI interface be equally accessible across all apps, so one could access it in Gmail, Docs, etc.,” Lazar said.

Lazar also compared Google’s Workspace Intelligence to Microsoft’s Work IQ, highlighting a shared ambition: an AI assistant that sees both personal activity and company‑wide knowledge. He warned that true usefulness will depend on the ability to ingest data from external systems such as CRM, ERP and project‑management tools, and on seamless operation in “mixed Google‑Microsoft‑Zoom‑Cisco‑Slack shops.” While third‑party integrations are arriving, Lazar remains uncertain about the timeline for full multi‑vendor support.

What to watch next

The rollout of Ask Gemini in Chat is just the first step. Future updates are expected to expand the roster of supported SaaS tools, tighten the privacy controls around Workspace Intelligence, and possibly expose the Gemini agent in other Google apps like Gmail and Calendar. Enterprises will be watching how quickly the AI can replace routine manual steps and whether the promised data isolation holds up under real‑world scrutiny. As the AI‑agent market matures, the balance between convenience and governance will likely shape adoption rates across both Google‑centric and mixed‑vendor environments.