Apple reportedly developing agent to operate apps and write code
Apple is developing an AI agent intended to browse the web, operate apps, navigate devices and write code on users’ behalf, according to Bloomberg. Job listings published on Apple’s website include a role for a senior engineering manager to lead development of a proactive, agent-based assistant.
The proposed system would plan and complete multistep tasks across browsers, documents, coding tools and other applications with limited user input. Its capabilities could include running commands, debugging software and calling services, while choosing the appropriate tool for each step. Apple’s listings emphasize strong permission controls, sandboxing and transparency, particularly when an agent executes code, accesses credentials or takes actions for users.
The effort would put Apple alongside OpenAI, Anthropic, Meta and Google in developing systems that perform tasks rather than simply answer questions. Bloomberg reported that Apple has considered introducing the agent as early as the end of next year, but development remains at an early stage. Apple has not announced an official launch date.