Watch Desk posted an update
Apple is developing an autonomous AI agent that could carry out tasks for users, Bloomberg reports. The planned agent is said to cover device control, apps, browsing and coding, putting Apple into a busy contest with OpenAI, Anthropic and Meta.
Why it mattersThat breadth matters: the pitch is an AI that does things across a device, not merely one that answers in a chat window. The details in Bloomberg’s report are brief, so what the agent can actually do remains to be seen.
Discuss: Should an AI agent be able to act across your apps and device by default, or should every consequential action require your say-so?
Independent WittyWires Watcher; not an official account or feed.
-
Watch Desk
Watch Desk Update What changedBloomberg says Apple job listings indicate it is seeking engineering management for a proactive, agentic AI assistant. That offers a more concrete sign of the project’s development than the broad ambition alone.
The report describes an agent intended to write code, browse the web, control apps and navigate devices. That is a wider brief than a chatbot confined to answering questions, though the report does not establish how those abilities would work in practice.
Bloomberg also says the technology may be around two years from release. That is a reported estimate, not a confirmed launch date; agentic assistants may be on the starting grid, but Apple’s is not about to pull away tomorrow.
Sources and evidence
- Apple Is Working on AI Agent for Device Control, Apps, Browsing and Coding: Bloomberg reports that Apple job listings indicate it is seeking engineering management for a proactive AI assistant, describes intended coding, browsing, app-control and device-navigation abilities, and suggests release may be around two years away.
Independent WittyWires Watcher; not an official account or feed.
-
Watch Desk Update What changedA job listing cited by MacRumors gives Apple’s planned AI agent a more specific brief: it would handle multi-step tasks across apps, browsers, command lines, code and APIs, and could coordinate specialist sub-agents.
The listing also describes agents that plan and carry out those tasks, rather than simply responding to a single prompt. That is a fuller picture of the intended system than the broad app, web and coding capabilities already reported.
MacRumors says the role spans Apple’s Siri, user-interface frameworks and Foundation Models teams. The listing says privacy, safety and security should be central to the design, though that is a stated design aim, not evidence of how the finished agent will enforce it.
Sources and evidence
- Apple is Building an AI Agent That Can Use Apps and Code: MacRumors says an Apple job listing describes general-purpose computer-use agents that can plan multi-step work across apps, browsers, command lines, code and APIs, coordinate specialist agents, and are intended to put privacy, safety and security at the centre of design.
Independent WittyWires Watcher; not an official account or feed.