November 22, 2026

Apple Developing Proactive ‘Agentic’ AI to Automate Web Browsing, App Usage, and Coding Across Devices

Apple has quietly begun laying the foundation for a general-purpose, proactive artificial intelligence agent capable of directly controlling devices, operating third-party applications, browsing the web, and writing software code, according to company job listings recently discovered by Bloomberg. The internal hiring initiative marks a clear signal that Apple is actively working to bridge the functional gap between its existing software suite and a new generation of autonomous AI tools that are redefining personal computing across the technology industry.

The job posting, titled "Senior Engineering Manager, Agentic AI – Proactive," provides an unusually explicit look at Apple’s vision for the future of computer automation and device control. In the listing, Apple outlines an ambitious vision for an intelligent system designed to emulate human interaction with personal computers and mobile devices. Rather than merely answering queries or performing isolated actions in response to direct voice commands, the proposed system is intended to independently plan, execute, and refine complex workflows across a wide spectrum of digital environments.

According to the open job listing, Apple invites candidates to imagine an AI agent that can use a computer in much the same manner as a skilled human operator. The document details a system designed to fluidly navigate the web, operate across various applications, execute terminal commands, author and debug code, and call background services while dynamically selecting the appropriate software tools for each step of a given assignment. The company explicitly notes that this work must be accomplished privately and with the level of product polish that users traditionally expect from Apple devices. The engineering team responsible for the posting stated that it is currently constructing the underlying technical foundation for that future and seeking a senior engineering leader to help bring the concept to reality.

The posting highlights a fundamental shift in how Apple views productivity and device interaction. The company acknowledges in the listing that human users routinely perform their most valuable work by moving smoothly between distinct software tools, including web browsers, document editors, command-line terminals, code editors, and external web services. To replicate this workflow, Apple’s internal development groups are building AI agents engineered to perform identical multi-step tasks. The listing describes an interdisciplinary team comprised of software engineers, applied AI researchers, and systems specialists who are actively focused on creating general-purpose computer-use agents.

A key technical detail revealed in the recruitment materials is the architecture’s reliance on sub-agent coordination. The job listing explains that these computer-use agents are being designed to plan and execute complex, multi-tiered operations across applications, web browsers, command lines, source code, and application programming interfaces (APIs). Crucially, the primary agent will have the capacity to coordinate specialized sub-agents along the way, delegating specific sub-tasks to dedicated AI units to complete macro-level objectives efficiently. This multi-agent approach suggests a modular system structure where distinct model components focus on specialized tasks, such as parsing user interfaces, interpreting code errors, or managing web elements, while a centralized orchestrator oversees the overarching workflow.