Google Labs is transforming its CC agent into a household tool with its own Google account, letting up to six family members share selected information so the agent can manage calendars, track tasks, and handle paperwork. The shift from a personal assistant to a collaborative family agent addresses a common friction point in busy homes: logistical details scattered across individual inboxes and mental checklists.
How the family agent works
CC now operates through a verified Google Account that gives it a distinct identity and a clear permissions model. Each family member chooses what to share - school emails, swim center schedules, vet appointments - and CC sorts through that shared context automatically. The agent creates a shared "Your Day Ahead" brief delivered each morning, showing who needs to be where and what tasks are pending or completed. It also connects to Calendar and Tasks, adding events and updating them as plans change.
"One of our top user requests was for CC to do more to help run a busy home," said Tom Shane, Senior Product Manager at Google Labs. "CC only sees what each member chooses to share."
Actions beyond reminders
The agent can complete time-consuming logistical work with user permission. It fills out permission slips and activity registration PDFs, builds school supply shopping lists, and crafts weekly meal plans. When it needs missing details, CC asks for them and updates its group memory. Behind the scenes, each CC runs on an isolated cloud computer powered by Google's agentic harness, Antigravity, and the latest Gemini models, allowing it to pre-fill forms, check live drive times via the Google Maps API, and create shared Docs or Sheets.
CC also has a shared memory that distinguishes household-level information - like a go-to grocery list or favorite family restaurants - from individual preferences such as dietary restrictions or time zones.
Privacy and control mechanics
Users can auto-share emails from specific senders like a child's school or travel bookings. Each week, CC provides a private list of new senders that members can choose to share. For one-off items, users can forward an email or send a photo of a birthday party invitation or soccer schedule via email or Google Chat, and CC will update the calendar. The agent only responds to group members and will not share information outside the group without permission.
This approach to shared automation reflects a broader trend in AI agents and automation, where systems move from single-user tools toward multi-person coordination. For professionals tracking product development and customer support workflows, the model of permissioned, memory-aware agents has direct parallels in enterprise environments where productivity gains depend on reducing coordination overhead.
Why this matters for customer support and product development
CC's architecture - a verified account, granular sharing controls, isolated cloud compute, and persistent group memory - mirrors design patterns emerging in enterprise AI agent deployments. Customer support teams evaluating agent-based tools should note the permission model: the agent acts only on explicitly shared data and asks before completing actions. For product developers, the shift from single-user to multi-user agent design signals where the market is heading. Household logistics may be the testbed, but the same coordination problems exist in team-based customer service and project workflows. The experiment is available now in the U.S. for users 18 and older with a personal Google account, with a waitlist for new users.
Your membership also unlocks:
