ChatGPT Group Chats Go Global, Bringing AI Into Team Conversations

OpenAI has begun a global rollout of ChatGPT group chats,  a shared conversation space for friends, families and teams that supports up to 20 participants, link-based invites, in-chat moderation and model routing via GPT-5.1 Auto. The move turns ChatGPT into a collaborative hub while raising fresh questions about UX, moderation and enterprise adoption.

Feature Breakdown: What’s Inside the Update

OpenAI has expanded its group-chat experiment into a global rollout that lets multiple people join a single ChatGPT conversation. Each group can host up to 20 members; creators share invite links, and new joiners see chat history when they enter. Core ChatGPT features, web search, file and image uploads, image generation and voice dictation — are available inside group chats. The feature is powered by GPT-5.1 Auto, which dynamically selects the model used to respond.

How it works

Users start a group from the people icon in the ChatGPT UI and invite participants via a link; first-time joiners create a brief profile (name, username, photo). ChatGPT behaves differently in these spaces — it “listens” and intervenes when helpful or when explicitly mentioned, rather than replying to every message. Memory does not carry over from private chats into groups: OpenAI says group conversations won’t populate a user’s private ChatGPT memory. Sensitive-content filters and age gating apply when minors are present. 

Collaboration, convenience and risk

Turning ChatGPT into a shared workspace changes the product’s role from a one-to-one assistant to a virtual facilitator. For teams, study groups and social planning, the feature reduces context switching: participants can co-create itineraries, budgets or draft documents in a single thread. Forbes and other outlets have flagged the potential productivity gains for workgroups and observed implications for workplace communication norms. At the same time, group AI introduces moderation and privacy trade-offs: who controls content, how model outputs are rate-limited, and how organisations adopt or restrict the feature.

For OpenAI, group chats are a logical extension of its consumer-to-collaboration playbook: they increase user engagement, expand product stickiness across plans (Free, Go, Plus, and Pro) and open avenues for integrations that surface in multi-user scenarios. The global rollout follows a short regional pilot and aligns with broader industry moves toward social/collaborative AI features. Enterprises will watch for admin controls, audit logs and data export policies before broad deployment. 

Key indicators to monitor in the coming months:

• User adoption and retention metrics for group chats versus one-to-one sessions; 
• moderation effectiveness and false-positive rates when the model auto-intervenes; 
• Enterprise administration features (SSO, discovery, export)
• Competitor responses — some platforms are already experimenting with multi-model, multi-participant threads.

How OpenAI balances interactivity with safety and privacy will shape whether group chats become a mainstream collaboration tool or remain a niche experiment