Computer use
An AI agent operating software the way a person would — looking at screens, clicking and typing — which gives it an execution surface when cleaner integrations or APIs are unavailable.
Computer use is a capability where an AI agent controls a computer through its visible interface: reading the screen, moving a pointer, clicking, typing and navigating applications. Instead of calling a purpose-built interface, it uses the one made for humans.
Its value is reach. Many systems have no API or a limited one, and computer use lets an agent work with them anyway. It is often paired with a cloud computer so the work can run unattended.
Trade-offs
It is a fallback, not the ideal. Operating a human interface is slower and more fragile than calling a defined action, and it hides ambiguity: the screen lets a person fill gaps that an agent may fill wrongly. A clean, described callable surface is more reliable and easier to permission. That is why capability readiness matters even when computer use works.
Read more in OpenAI Dots is a test of whether AI can carry a goal, not just complete a task.
Related terms
Cloud computer
A remote machine an AI agent can use to run tools and applications for long tasks, so work continues without the user's own device. It removes one constraint and exposes others, like permissions and oversight.
Agentic AI
AI systems that pursue a goal by planning, using tools and taking multi-step actions with some autonomy, rather than only answering a single prompt.
Callable surface
The set of actions an organization exposes so AI agents can invoke them directly, each described precisely enough to be agent-ready: meaning, inputs, permissions, side effects and recovery.
Used in these essays
OpenAI Dots is a test of whether AI can carry a goal, not just complete a task
Dots matters less as another capable assistant than as a test of persistent delegation: can AI keep carrying a goal without giving the user a new system to manage?
The next AI interface may never be seen
Agents are turning software capabilities into an interface of their own. The next enterprise design problem is deciding what should be callable, by whom, and under which boundaries.