Computer use
Also called computer-use agent, computer-using agent
An AI that operates a computer the way you do: it looks at the screen, moves the pointer, clicks and types to finish a task.
Most AI tools work through connections: a connector to your email, a plugin for a booking site. Computer use skips that. The model takes a screenshot, works out what's on it, decides where to click or what to type, does it, and takes another screenshot to see what happened. Repeat until done.
That means it can work in almost any app, including old ones with no connection to anything. It also means it's slow, makes mistakes a person wouldn't, and can be misled by what it reads on screen: a message or web page can contain instructions it might follow.
So the apps that offer it ask permission per app or site and pause before important steps. Those pauses are the safety feature; don't switch them off for convenience.
For example
You ask it to copy the totals from 12 PDF invoices into a spreadsheet, and watch it open each one and type the numbers in.
Put it to use
Where you'll meet computer use in ChatGPT, Gemini and Claude, how to use it there, and the mix-up most people make.
Daily Bytes Plus is in the Android app.