ChatGPT Group Chats Go Global With New Privacy Rules and Collaborative Tools

OpenAI has expanded ChatGPT with a global release of group chats, giving users the ability to bring the assistant into conversations with up to twenty participants. The feature is available through the ChatGPT app and the web, following a smaller pilot. It is designed to let the AI take part without dominating the discussion. ChatGPT remains passive until tagged or until the context makes participation useful, which reduces the risk of the assistant interrupting or steering conversations unnecessarily.

The system separates group chats from private sessions. Nothing from an individual user’s memory transfers into a group, and information shared inside a group does not enter ChatGPT’s long term memory. The assistant also encourages users to set a simple profile so they avoid appearing as unlabelled icons. Additional safeguards activate automatically when children join, tightening content filters for every participant. The group creator can leave at any time, but they remain the only person who cannot be removed by others.

 

 

Early tests show that ChatGPT’s behaviour adapts to the conversation type. In discussions where participants were comparing local breakfast spots, the AI offered details when asked but remained mostly in the background. In a simulated debate over choosing a movie, ChatGPT helped narrow options by referencing earlier comments about genre preferences. It handled side conversations and returned to the main topic when prompted, acting as a neutral participant that helped keep the discussion organised.

The assistant becomes more useful when disagreements arise or when groups struggle to reach clear decisions. It can provide factual context, suggest options and highlight preferences already expressed by the group. The feature depends on users tagging the AI at the right moments rather than expecting it to moderate conversations automatically. This design choice prevents group chats from turning into AI led sessions.

ChatGPT can respond with emojis, reference user profile pictures and create personalised images featuring members of the group when asked. These touches make the assistant feel more integrated with the conversation. The system uses GPT 5.1 Auto, allowing it to adjust its explanations based on the group’s needs and the complexity of the task.

The introduction of group chats does not expand ChatGPT’s core capabilities, but it changes how people might use the assistant. The presence of an AI that can summarise, recall earlier points and provide neutral suggestions gives groups more momentum when planning activities or resolving choices. Coordinating with several people often becomes inefficient, and a steady assistant that tracks preferences and provides structured information can reduce friction.

As group communication continues to rely on messaging apps and shared channels, adding a responsive AI that contributes when needed may become a standard part of collaborative work. ChatGPT’s new group chat feature is an early step toward that direction, offering a structured way for groups to make decisions and share information without losing control of the conversation.