Apple is reportedly exploring an AI agent designed to perform computer tasks on behalf of users, according to Bloomberg. The company posted a job listing for a senior engineering manager to lead work on an “AI agent that can use a computer,” describing software that can navigate websites, operate apps, run commands, write and debug code, and call services.
Capabilities described in the job listing
The posting, cited by Bloomberg, outlined an agent that chooses appropriate tools for each step and moves fluidly between apps, browsers, command lines, code and APIs. It emphasised privacy and a polished user experience as development goals. Bloomberg reported the listing framed the agent as able to carry out multi-step tasks across different environments on the machine.
Context and platform controls
Agents that can perform computer work have become an active area of development across the industry. Bloomberg noted that major labs such as Anthropic and OpenAI have launched similar services, and that Meta recently introduced its Muse agent line.
At the same time, Bloomberg reported Apple is tightening controls on third-party agents. The outlet says macOS 27.3 will deploy warnings discouraging users from granting full access to external agents. In October, Apple notified developers it planned to introduce additional controls to ensure that apps requiring extraordinary access can obtain it only after very explicit user action.
Bloomberg added that the in-house agent project is in early stages. The report said Apple has eyed introducing the service as soon as the end of next year, but that progress remains preliminary.
The reporting preserves the company names, the job title cited and the specific platform references Bloomberg included: macOS 27.3, Anthropic, OpenAI, Meta and Muse. Bloomberg is the outlet that published the job-listing details and the notes about platform changes and timing.
Original source: TechCrunch AI