Anthropic unveiling a new AI tool that can analyze your computer screen and take over to act on your behalf.

The latest in an escalating battle for AI-powered agents that can help automate tasks.

Are Deirdre Bosa is watching that in today's Tech Jack.

Good morning, Dee.

Hey, good morning, Carl.

So a new race is off, the 01 model from Open AI

earlier this year, now Anthropics new feature,

which are called computer use that's giving these models new

abilities and users new ways to use AI potentially putting Google on the

backfoot which has been working to catch up and release its own next generation AI model that can reason and better

automate tasks.

Now yesterday guys we talked about agents as the next frontier in AI, the enterprise

race to put agentic products into their customers hands but with

computer use from Anthropic that now lets users use AI to control actions on

their computer like clicking buttons and typing to browse the internet or make

calendar appointments.

Its model Claude can essentially, the company says, carry out actions on the user's behalf

even without their consent. so acting like a human

collaborator or agent that word again now as we've come to expect with new AI

products Anthropic says big disclaimer here that the technology is still

experimental and at times cumbersome and error prone.

But obviously the company thought it worth the risk to be the first it says frontier AI model

...