Editorial illustration for Apple seeks engineers to build AI agent that browses web and writes code
Apple Hires for AI Agent That Codes and Browses Web
Apple seeks engineers to build AI agent that browses web and writes code
Apple posted a job listing looking for an engineer to lead work on an agentic AI assistant, one that can browse the web, move across apps, write code, and call outside services on a user's behalf. Bloomberg spotted the posting, which places Apple squarely in a race it's been slow to join. Meta has Muse, OpenAI has Dots, Anthropic has Duet, and Google has Tasks, all variations on the same idea: software that doesn't just answer questions but actually does the clicking, typing, and navigating for you. Some of these tools already run through Mac apps, taking control of a computer directly, and Google recently gave Gemini that same ability.
Apple's version would build on the AI-upgraded Siri it started rolling out in September, which can already read what's on a user's screen and act within apps. The new hire would reportedly work across Apple's Siri, UI Frameworks, and Apple Foundation Models teams, suggesting the agent isn't a side project but something tied directly into how Siri evolves next. The listing itself lays out the ambition plainly.
Apple is working on its own AI agent that can perform a range of tasks for you, including browsing the web, working across apps, writing code, and calling services. That’s according to a job listing spotted by Bloomberg, which says Apple is looking for an engineer to lead the creation of an agentic AI assistant.
Why this matters
Apple is late to agentic AI, and this job listing confirms it. Google has Project Mariner, OpenAI has Operator, and Anthropic's Claude already calls tools and browses on command. Apple, by contrast, is still at the hiring stage for an engineer to "lead the creation" of something comparable.
For developers watching the agent space, that gap matters less than what Apple might do with it once built. Apple controls the hardware, the OS, and the app ecosystem on over a billion devices. An agent that can browse, write code, and call services while sitting inside iOS or macOS has a distribution advantage no startup can match, regardless of how far behind it started.
For founders building agent products right now, this is a signal to move fast and build something sticky before Apple ships its version. For researchers, the real question is what "working across apps" actually means on a platform built around sandboxing and privacy restrictions. Apple's walled garden has made third-party automation hard for years. Whether that same architecture lets its own agent do things others can't, or forces it into the same limits, is the detail worth watching once more specifics leak.
Common Questions Answered
What specific capabilities is Apple's agentic AI assistant expected to have?
According to the job listing, Apple's agentic AI assistant will be capable of browsing the web, moving across multiple apps, writing code, and calling outside services on a user's behalf. These capabilities allow the AI to perform complex tasks autonomously rather than simply answering questions.
How does Apple's position in agentic AI compare to competitors like Google, OpenAI, and Anthropic?
Apple is significantly behind competitors in the agentic AI race, currently at the hiring stage for an engineer to lead development of their assistant. In contrast, Google has Project Mariner, OpenAI has Operator, and Anthropic's Claude already has the ability to call tools and browse on command, demonstrating that these companies are further along in deployment.
Why does Apple's entry into agentic AI matter despite being late to the market?
Apple's potential advantage lies in its control over hardware, operating systems, and the app ecosystem across over a billion devices worldwide. Once Apple develops a comparable agentic AI assistant, this unique position could allow it to integrate and deploy the technology at a scale that competitors cannot match.
What other companies have developed agentic AI assistants similar to what Apple is building?
Meta has created Muse, OpenAI has developed Dots, Anthropic has built Duet, and Google has created Tasks, all representing variations on the same concept of agentic AI. These assistants share the core functionality of performing clicking, typing, and navigating tasks rather than just providing answers to user questions.
Further Reading
- Apple is hiring engineers to build its own AI agent - The Verge
- Apple Is Working on AI Agent for Device Control, Apps, Browsing and Coding - Bloomberg
- Apple Is Building an AI Agent That Can Use Apps and Code - MacRumors
- Report: Apple job listings detail plans for autonomous AI agents - 9to5Mac
- Apple job listing gives away proactive agentic AI project - Macworld