It was initially present in leaked code and publicized by AI influencers on X, however OpenAI has made it official: ChatGPT now affords Group Chats, permitting a number of customers to hitch the identical, single ChatGPT dialog and ship messages to one another and the underlying giant language mannequin (LLM), on-line and by way of its cell apps.
Think about including ChatGPT as one other member of your present group chats, permitting you to textual content it as you’ll certainly one of your pals or members of the family and have them reply as properly, and also you'll have an thought of the intriguing energy and potential of this characteristic.
Nonetheless, the characteristic is simply obtainable as a restricted pilot for now to ChatGPT customers in Japan, New Zealand, South Korea, and Taiwan (all tiers, together with free utilization).
“Group chats are only the start of ChatGPT changing into a shared house to collaborate and work together with others,” OpenAI wrote in its announcement.
This growth builds on inside experimentation at OpenAI, the place technical staffer Keyan Zhang mentioned in a publish on X that OpenAI's crew initially thought-about multiplayer ChatGPT to be “a wild, out-of-distribution thought.”
In line with Zhang, the mannequin’s efficiency in these early checks demonstrated way more potential than present interfaces sometimes permit.
The transfer follows OpenAI investor but competitor Microsoft's replace of its Copilot AI assistant to permit group chats final month, in addition to Anthropic's introduction of shareable context and chat histories from its Claude AI fashions by its Initiatives characteristic launched summer time 2024, although this isn’t a simultaneous, realtime group chat in the identical method.
Collaborative performance built-in into ChatGPT
Group chats perform as shared conversational areas the place customers can plan occasions, brainstorm concepts, or collaborate on tasks with the added help of ChatGPT.
These conversations are distinct from particular person chats and are excluded from ChatGPT’s reminiscence system—that means no information from these group threads is used to coach or personalize future interactions.
Customers can provoke a bunch chat by deciding on the folks icon in a brand new or present dialog. Including others creates a duplicate of the unique thread, preserving the supply dialogue. Individuals can be a part of by way of a shareable hyperlink and are prompted to create a profile with a reputation, username, and photograph. The characteristic helps 1 to twenty individuals per group.
Every group chat is listed in a brand new part of the ChatGPT interface, and customers can handle settings like naming the group, including or eradicating individuals, or muting notifications.
Powered by GPT-5.1 with expanded instruments
The brand new group chat characteristic runs on GPT-5.1 Auto, a backend setting that chooses the optimum mannequin based mostly on the person’s subscription tier and the immediate.
Performance equivalent to search, picture era, file add, and dictation is accessible inside group conversations.
Importantly, the system applies fee limits solely when ChatGPT is producing responses. Direct messages between human customers within the group don’t rely towards any plan’s message cap.
OpenAI has added new social options to ChatGPT in help of this group dynamic. The mannequin can react with emojis, interpret conversational context to resolve when to reply, and personalize generated content material utilizing members’ profile pictures—equivalent to inserting person likenesses into photos when requested.
Privateness by default, controls for youthful customers
OpenAI emphasised that privateness and person management are integral to group chat design. The characteristic operates independently of the person’s customized ChatGPT reminiscence, and no new recollections are created from these interactions.
Participation requires an invite hyperlink, and members are at all times in a position to see who’s in a chat or go away at any time.
Customers underneath the age of 18 are routinely shielded from delicate content material in group chats. Mother and father or guardians can disable group chat entry altogether by way of built-in parental controls.
Group creators retain particular permissions, together with immunity from being eliminated by others. All different individuals will be added or eliminated by group members.
A testbed for shared AI experiences
OpenAI frames group chats as an early step towards richer, multi-user functions of AI, hinting at broader ambitions for ChatGPT as a shared workspace. The corporate expects to develop entry over time and refine the characteristic based mostly on how early customers interact with it.
Keyan Zhang’s publish means that the underlying mannequin capabilities are far forward of the interfaces customers at present work together with. This pilot, in OpenAI’s view, affords a brand new “container” the place extra of the mannequin’s latent capability will be surfaced.
“Our fashions have much more room to shine than right this moment’s experiences present, and the present containers solely use a fraction of their capabilities,” Zhang mentioned.
With this preliminary pilot centered on a restricted set of markets, OpenAI is probably going monitoring each utilization patterns and cultural match because it plans for broader deployment. For now, the group chat experiment affords a brand new method for customers to work together with ChatGPT—and with one another—in actual time, utilizing a conversational interface that blends productiveness and personalization.
Developer entry: Nonetheless unclear
OpenAI has not offered any indication that Group Chats will probably be accessible by way of the API or SDK. The present rollout is framed strictly inside the ChatGPT product setting, with no point out of device calls, developer hooks, or integration help for programmatic use. This absence of signaling leaves it unclear whether or not the corporate views group interplay as a future developer primitive or as a contained UX characteristic for finish customers solely.
For enterprise groups exploring easy methods to replicate multi-user collaboration with generative fashions, any present implementation would require customized orchestration—equivalent to managing multi-party context and prompts throughout separate API calls, and dealing with session state and response merging externally. Till OpenAI gives formal help, Group Chats stay a closed interface characteristic slightly than a developer-accessible functionality.
Here’s a standalone concluding subsection tailor-made for the article, specializing in what the ChatGPT Group Chat rollout means for enterprise choice makers in each pilot areas and globally:
Implications for enterprise AI and information leaders
For enterprise groups already leveraging AI platforms—or getting ready to—OpenAI’s group chat characteristic introduces a brand new layer of multi-user collaboration that would shift how generative fashions are deployed throughout workflows. Whereas the pilot is restricted to customers in Japan, New Zealand, South Korea, and Taiwan, its design and roadmap supply key indicators for AI engineers, orchestration specialists, and information leads globally.
AI engineers managing giant language mannequin (LLM) deployments can now start to conceptualize real-time, multi-user interfaces not simply as help instruments, however as collaborative environments for analysis, content material era, and ideation. This provides one other entrance in mannequin tuning: not simply how fashions reply to people, however how they behave in dwell group settings with context shifts and diversified person intentions.
For AI orchestration leads, the power to combine ChatGPT into collaborative flows with out exposing non-public reminiscence or requiring customized builds might cut back friction in piloting generative AI in cross-functional groups. These group classes might function light-weight alternate options to inside instruments for brainstorming, prototyping, or information sharing—helpful for groups constrained by infrastructure, funds, or time.
Enterprise information managers might also discover use instances in structured group chat classes for information annotation, taxonomy validation, or inside coaching help. The system’s lack of reminiscence persistence provides a stage of information isolation that aligns with commonplace safety and compliance practices—although world rollout will probably be key to validating regional information dealing with requirements.
As group chat capabilities evolve, choice makers ought to monitor how shared utilization patterns may inform future mannequin behaviors, auditing wants, and governance constructions. In the long run, options like these will affect not simply how organizations work together with generative AI, however how they design team-level interfaces round it.
