OpenAI is testing a beta feature that could make ChatGPT feel less like a single chatbot and more like a coordinated workspace for specialized assistants. The feature, described as multi-GPT conversations, lets users call different GPTs into the same chat window by typing "@" followed by the name of a GPT.
On its own, that may sound like a small interface change. But the bigger idea is clear: ChatGPT could become a place where several custom GPTs work inside one conversation, with their answers and abilities connected through the same thread.
How multi-GPT conversations work
The feature allows a user to mention a GPT directly inside a chat. Instead of leaving one conversation and starting another, the user can summon a specific GPT into the existing exchange.
That means a user could ask one GPT for an answer, then bring in another GPT to respond to it, compare it, or build on it. The source article describes this through two example chatbots: "Talk with Donald - 2024" and "Biden Simulator". Both are presented as chatbots that speak in the style of the two politicians.
In the test described, the user first asked the Trump chatbot for its opinion on Joe Biden. The user then asked the Biden chatbot, which supposedly responds in the style of Biden's speeches, to react to Trump's statement. The important technical point is that the chatbots could explicitly refer to each other inside the same conversation.
For everyday use, that structure could matter more than the specific example. A single chat window could become a shared context where specialized GPTs are invited as needed, instead of being isolated tools with separate histories.
Why @GPT matters for OpenAI's assistant plan
The source frames multi-GPT conversations as part of OpenAI's broader ambition to make ChatGPT a personalized, individual and universal assistant. The feature does not fully deliver that vision by itself, but it creates a visible path toward it.
Today, the user chooses which GPT to call by typing "@" and the GPT's name. The next step, according to the source, would be for the underlying model to automatically learn which GPT should respond to which request.
That would shift ChatGPT from a tool the user manually directs into a system that can route tasks across different capabilities. In plain terms, the assistant would not only answer questions. It would also decide which specialized GPT is best suited to help.
Sam Altman connected this broader direction to personalization in a recent podcast with Bill Gates. He said that "customizability and personalization " are key items on OpenAI's development roadmap.
People want very different things out of GPT-4: different styles, different sets of assumptions. We’ll make all that possible, and then also the ability to have it use your own data. The ability to know about you, your email, your calendar, how you like appointments booked. Those will be some of the most important areas of improvement.
That statement helps explain why a feature like @GPT is more than a convenience shortcut. If ChatGPT is meant to adapt to different styles, assumptions, data, and personal workflows, then a multi-GPT environment gives OpenAI a way to combine many specialized behaviors in one place.
The personalization promise
The appeal of a universal assistant is not only that it can do more. It is that it can respond in ways that match the user's habits and preferences.
The source points to several areas where OpenAI sees improvement: different styles, different sets of assumptions, the ability to use a person's own data, and knowledge of email, calendar, and appointment preferences. Those examples show a shift from generic answers toward context-aware help.
Multi-GPT conversations fit that direction because they make specialization more flexible. One GPT might have a particular tone or task focus, while another might approach the same conversation differently. In a shared chat, those differences can become part of a single workflow rather than separate sessions.
For users, the practical value would be continuity. The conversation can keep its context while the user brings in another GPT with a different role or skill. That is a step toward the kind of assistant that understands not only the current prompt, but also how several tools should work together around it.
Political GPTs raise a difficult question
The source also highlights a tension inside the GPT Store. Political chatbots remain available there, including the Donald Trump chatbot used in the example.
According to its creator, the Donald Trump chatbot has been trained with Trump's books, speeches, and debates. It responds in Trump's style and represents at least some of his political views.
The issue is that OpenAI has banned the use of its technology for political propaganda, according to the source. An outside political chatbot that mimicked Democratic candidate Dean Phillips using OpenAI's models was banned by the company under those rules.
The source article says OpenAI did not respond to three requests for comment about why the political chatbots are allowed in the GPT Store. It also says the political GPTs remained available without any changes or restrictions, suggesting that the company doesn't care or doesn't have the capacity to address these issues.
That unresolved enforcement question sits beside the product story. Multi-GPT conversations could make ChatGPT more useful and more personal. But as GPTs become easier to combine in one chat, OpenAI will also face sharper questions about which GPTs belong in that ecosystem and how consistently its rules are applied.
What changes next
The @GPT feature is still described as a beta test, but it shows the direction of travel. OpenAI is making ChatGPT less like a single assistant and more like a platform for many assistants that can share a conversation.
If the system eventually learns which GPT should answer which request, users may spend less time managing tools and more time working through tasks. That is the central promise of a universal assistant: not just more chatbots, but a single place where the right kind of help appears at the right moment.
The open question is governance. The same feature that makes personalized AI assistants more powerful can also make controversial or sensitive GPTs easier to use together. OpenAI's next challenge is not only technical. It is also about how the company manages the GPT Store as ChatGPT becomes a more universal layer for everyday life.