- A job listing shows Apple is hiring a senior engineer to lead a new on-device AI agent, positioning it against Meta’s Muse, OpenAI’s Dots, and xAI’s Grok Bots.
- Bloomberg reports the agent could arrive as early as late 2027 and remains in an early stage.
- Apple CEO John Ternus is reportedly pushing software chief Craig Federighi toward a bolder AI strategy.
- macOS 27.3 now requires 14 steps, up from four, before a third-party agent like ChatGPT can control a Mac.
What Happened
A job listing reviewed by Bloomberg shows Apple is hiring a senior engineer to lead development of an AI agent capable of controlling a user’s device, joining Meta, OpenAI, Anthropic, and Google in the category.
Why It Matters
Apple has moved cautiously on generative AI, rolling out a revamped, Dots-style Siri agent only after internal rifts and staff departures, per the report. A device-controlling agent would mark a shift toward the agentic AI race already underway at rivals like Meta’s Muse.
Technical Details
Privacy is expected to be central to the design, with permission models and sandboxed test environments limiting what the agent can access, according to Bloomberg. The project remains early-stage, with a possible late-2027 arrival. Separately, Apple’s macOS 27.3 update raised the steps required for a third-party agent like ChatGPT to control a Mac from four to 14.
Who’s Affected
Apple CEO John Ternus is reportedly pushing software chief Craig Federighi toward a more aggressive AI roadmap. Third-party AI companies that rely on cross-app device control say the macOS permission change could slow their products, though Apple frames it as a security measure, per Bloomberg.
What’s Next
With no confirmed release date, Apple’s agent would arrive well behind Meta, OpenAI, Anthropic, and Google’s existing device-control agents, leaving open whether tighter permissions on its own platforms give it an edge once it ships.