Realtime AI News
Apple reportedly building an AI agent that can write code and operate apps
Reports on November 20 said Apple is developing an AI agent capable of writing code and using apps to carry out tasks, putting it in the same category as agent products from rival AI companies. The coverage was also read as a signal that Apple does not intend to sit out the AI race, though no timeline, model source or product form has been disclosed.
Outlets including GSMArena and Yahoo Finance reported on November 20 that Apple is developing an AI agent that can write code and use apps to complete tasks. The wording is hedged, and Apple has not confirmed the product or described it publicly.
As described, the agent belongs to the same family as the coding assistants and computer-use agents rivals have shipped over the past year: software that executes multi-step work rather than answering questions. If that capability lands at the system level, Siri's role shifts from answering to acting, which is where Apple has its strongest opening in the agent race.
Apple is not starting from zero. Through mechanisms such as App Intents, Siri and Apple Intelligence have gradually gained reach into functions inside apps without opening them. An agent is the logical next step; the hard part is going from invoking a single function to reliably chaining a sequence of them.
Competitive pressure is why the report travelled. OpenAI, Google and Anthropic have turned coding agents and computer-use into shipping products and are pushing into enterprises. Apple's leverage is different: the device install base, system-level permissions and on-device processing, which support a privacy story that cloud-first rivals cannot easily match.
What the public record does not contain is nearly everything else. Whether the model is in-house or licensed, whether inference runs on device, whether developers get an interface first, and how the work lines up with the next OS cycle are all unstated. Yahoo Finance's framing, that John Ternus is done sitting out the AI race, is interpretation layered on top of the same report.
For developers, the interface matters more than the rumour. If Apple eventually exposes system-level agent capability, app makers will need to decide which actions may be delegated, which data stays on device and how permissions are audited. Apple has historically solved such problems with controlled frameworks, which makes a general-purpose operate-anything agent unlikely.
Three signals are worth watching: whether the next OS cycle comes with a stated timeline, whether developer documentation gains an agent-related framework, and whether Apple's language on privacy and local inference shifts to match. Until then, the value of the report is directional, since it confirms Apple is filling in the agent gap.
Why it matters
If the report holds, agentic control of iOS and macOS would reshape how users reach third-party apps and how permission models work. Until Apple ships or announces something, the timeline, model sourcing and openness all remain open questions.
Nearby Updates
All11/21, 01:03
Signal blocks screenshots by default to keep OS-level AI agents out of chats
On November 20 Signal said it will block in-app screenshots by default on iOS 27 and enable screen protection on Android when an AI assistant tries to read the app, arguing that OS-level agents capture chat content and hand it to local or remote databases. Signal concedes the remedy is incomplete and unimplementable on macOS, and calls on Apple, Google and Microsoft to give developers tools to opt out of AI data access.
11/21, 01:49
Arena Revises Alignment Index Method, OpenAI Takes All Seven Top Spots
Arena updated the methodology behind its Alignment Index on November 20, saying the revised signals catch model failures more broadly and reliably, with OpenAI models holding all seven top positions. The index scores models on unauthorized actions, false attribution and deceptive completion seen in real agent sessions, making such rankings highly sensitive to how failures are defined.
11/21, 01:53
Google's Gemini clears the static from firefighter radio, cutting Watch Duty alerts to under 30 seconds
Google detailed on November 20 how a Gemini-powered tool built with the nonprofit wildfire app Watch Duty transcribes noisy fire radio into clean, searchable text and flags new dispatches in under 30 seconds. Word error rate on degraded VHF audio fell from 38% to 22.34%, and volunteer reporter satisfaction with transcription rose from 35.3% to 92.3%.
11/21, 02:15
METR's updated time horizon plot: exponential gains continue, judgment lags
METR has published an updated version of its widely cited and widely misread time horizon plot, adding longer and more open-ended tasks to probe how models handle work that cannot be scored automatically. The report finds capabilities on verifiable tasks still improving exponentially, while models remain below human level on judgment-heavy work such as designing evaluations or scoring philosophical arguments.