If you use a Mac alongside an automated assistant, your daily setup will soon shift. Apple wants to take over the digital chores you currently give to outside software, but its own assistant remains in early development.
Apple Inc. is developing its own autonomous artificial intelligence assistant that can get things done on behalf of users. The new Apple AI agent project surfaced this week in job postings placed on the company website. The effort puts Apple against OpenAI, Anthropic PBC, Meta Platforms Inc. and others. Bloomberg News reported earlier this month that Apple was developing an agent offering, but new listings provide fresh evidence of the plan. Apple has eyed introducing the service as soon as the end of next year, though development remains in the early stages. An Apple spokesperson did not respond to a request for comment.
For Mac owners who rely on third-party tools, the shift could alter daily computing habits. Apple says safety must be built into the system’s design, and the listing calls for strong permission models, sandboxing and transparency. Users who prefer independent software may encounter more friction during daily setup. The listing describes agents that can plan and carry out multistep tasks across apps, browsers, command lines, code and application programming interfaces.
Inside Apple’s Agent Plans
In one listing, Apple seeks a senior engineering manager for a proactive, agentic AI assistant. The system will write code, browse the web, control apps and operate devices on a user’s behalf, according to the post. Apple’s new chief executive officer, John Ternus, has pushed software engineering chief Craig Federighi toward a bolder AI approach. An agent remains key to that effort.
The listing states: “Imagine an AI agent that can use a computer the way a capable person does. It browses the web, works across apps, runs commands, writes and debugs code, and calls services, and it chooses the right tool for each step. It does all this privately and with the polish people expect from Apple.”
The listing says a team is building the “foundation for that future” and seeks the manager to “make it real.” It states: “People already do their most valuable work by moving fluidly between tools: a browser, a document, a terminal, an editor, a service. We’re building AI agents that can do the same.” The post adds Apple is developing agents that can “plan and carry out multistep tasks” across apps, browsers, command lines, code, and application programming interfaces, and can “coordinate specialized subagents along the way.”
The listing cites “general-purpose computer-use agents,” technology allowing ChatGPT and Muse to control a Mac. Google Gemini added that functionality earlier on Thursday.
Rivals Across the Mac Ecosystem
The move follows the breakout success of Meta’s Muse, which can make restaurant reservations, book travel and prepare news briefings. Apple’s App Store has been dominated by AI agents. Muse long held the No. 2 spot among free apps on Thursday, behind TikTok. OpenAI’s ChatGPT and Anthropic’s Claude were also in the top five.
Anthropic recently widened testing of its agent, Duet, while OpenAI expanded Dots. Alphabet Inc.’s Google overhauled its agent, now called Tasks, earlier this week. Like Elon Musk’s Grok Bot, Tasks lets the system make phone calls on a user’s behalf. OpenAI, Anthropic and Musk’s SpaceXSI are behind developer coding tools Codex, Claude Code and Cursor.
Shares of Meta and Anthropic slid in late trading after Bloomberg News reported on Apple’s efforts. Meta fell nearly 2 per cent, while Anthropic dropped more than 3 per cent.
Privacy Design and Platform Restrictions
Privacy is central, with Apple saying safety must be built into system design. “Agents that can run code, use credentials, and act on a user’s behalf need strong permission models, sandboxing and transparency,” the listing states.
In the beta of forthcoming macOS 27.3, Apple introduced warning screens prompting users to reconsider granting third-party agents permission to control their Macs. With this change, ChatGPT users must complete 14 distinct steps to enable third-party computer-use tools, up from four.
Apple positioned the shift as protecting customers, though some AI firms and users view it as an effort to slow competitors. Apple has already been sued by the US Justice Department for actions such as blocking third-party access to features and hardware.
Apple is tightening the door on OpenAI and Anthropic’s Mac agents at the exact moment it hires the team that will walk through that entrance. The forthcoming macOS 27.3 beta adds warning screens to third-party computer-use tools. That same software release prepares the desktop environment for Apple’s own agent arrival.
Is Apple building an AI agent?
Yes. Apple is developing an autonomous AI agent that can use a computer the way a person does. Job listings describe a system that browses the web, works across apps, runs commands, writes and debugs code, and calls services. Apple has considered launching it as soon as late 2027, but the project is at an early stage.
Apple is moving forward with its own autonomous assistant to manage tasks across desktop computers. Rival firms lead the app marketplace, while upcoming operating system updates alter how external tools function on Macs. The key follow-up trigger will come when Apple updates its job boards with more engineering openings or shares public details about its software schedule.



