AI assistants may be moving beyond answering questions. OpenAI’s Dots are designed to keep working toward a goal in the background, connect with everyday apps and use their own cloud computers. That shift raises a practical question: how much independence will people be comfortable giving a personal AI agent?
From chatbot responses to ongoing tasks
Dots are presented as a step beyond the familiar prompt-and-response model. Rather than simply waiting for a person to ask another question, they can continue working toward a goal in the background. Their connections to commonly used apps could also give them access to more of a person’s digital life.
The product’s cloud computers are part of that design. Together, the ability to work in the background and connect to apps suggests a system that can take on continuing tasks, rather than only provide a reply in a chat. The source does not specify which tasks Dots will handle, so the central point is the broader change in how an AI product could operate.
Autonomy brings questions of control
Giving an agent room to act can make it more useful, but it also changes what a person needs to decide. Users may want to know which actions an agent can take independently, when it should stop and ask, and how much authority it should have over connected apps.
Those decisions become more significant as systems gain access to sensitive information. A personal agent that works across everyday apps may need to understand more about a user’s work and digital life. The more access it has, the more important it becomes for people to feel confident about what the agent can see and do.
That makes trust a product question as much as a privacy question. People may weigh the convenience of an agent that continues working against the control they give up. Whether users are willing to make that trade depends in part on whether companies can earn enough confidence for people to use these systems.
Embiricos to discuss what comes next
Alexander Embiricos, OpenAI’s product lead for Dots, is scheduled to speak with TechCrunch at TechCrunch Disrupt 2026. His session, “What Comes After the Chatbot?”, will cover the thinking behind Dots, OpenAI’s view of personal agents and the role privacy and trust could play in whether they succeed.
The session is set for the event’s AI Stage. Disrupt 2026 takes place October 13-15 at Moscone West in San Francisco. Embiricos will appear days after OpenAI unveiled its new always-on personal AI agent, putting the conversation close to the product’s introduction.
Why the next step depends on confidence
The promise of an agent is that it can keep moving a task forward without needing a person to restart the conversation each time. But that convenience only matters if users are comfortable with the agent’s reach and decisions. More persistent work and wider app connections create a need for clear expectations about access and action.
Dots therefore point to a larger question for personal AI: what should these systems be allowed to do on a person’s behalf? The answer will shape how people use them, especially when an agent is working beyond the visible exchange of a chat. At Disrupt, Embiricos is expected to address how OpenAI sees that future and why trust may determine whether it takes hold.