Instinct Expands AI Agent to Group Conversations, Allowing Non-Users to Participate
Instinct, a $10 billion-valued AI agent startup, is launching group chat functionality that enables collaborative use of its AI assistant for tasks like travel planning and event coordination, even among friends who haven't created their own Instinct accounts. The feature implements privacy controls where personal AI agents request permission before sharing information with group instances, and group data remains siloed from individual user accounts. The rollout begins with early access users and positions Instinct competitively against rivals like Meta's Muse in the consumer AI agent market.
Instinct's $10 billion valuation reflects investor confidence in the consumer AI agent market, where the startup now competes directly with Meta's Muse and OpenAI's recently launched ChatGPT Dots. The group chat feature addresses a common friction point in social coordination—the scattered decision-making that typically occurs across multiple message threads when friends plan activities together. By allowing non-account holders to participate, Instinct removes a barrier to adoption that might otherwise limit the utility of the tool for group scenarios.
The technical architecture isolates group instances from personal user data, with explicit permission layers built into the system. When new participants join a group, the system holds back pending responses from personal agents until those responses can be reviewed, adding another safeguard to the data-sharing process.
The expansion into group collaboration could reshape how friends organize shared activities, potentially reducing time spent coordinating logistics. However, the move also introduces new data-sharing dynamics that may raise privacy considerations as group instances accumulate information about multiple users' preferences and decisions. The competitive pressure among AI agent platforms to add collaborative features could accelerate adoption, though the long-term impact will partly depend on whether users find the privacy-control mechanisms sufficiently transparent and trustworthy in practice.