Zendoric
← Back to the day · November 12, 2026

Apple's agentic Siri push is a catch-up race in which control of the platform matters more than model quality

🕒 Published on Zendoric: November 12, 2026 · 00:05

✨ AI-generated · how it's made

According to Bloomberg's Mark Gurman, CEO John Ternus is pushing Craig Federighi to make Siri a real agent. First comes control of Apple's own apps, then third-party apps in the spring, with ongoing multi-step tasks possibly next year. Apple is behind on models, but it owns the device everything runs on.

Apple wants Siri to become an AI agent: a system that takes actions for you instead of just answering. Bloomberg's Mark Gurman reports that CEO John Ternus is pressing AI chief Craig Federighi to speed up the Siri roadmap. The goal is a product that can compete with OpenAI's Dots and Meta's Muse. Gurman says Federighi delivered a working Siri AI on a very short timeline. He also says Apple knows it is not keeping pace with its rivals.

The roadmap has three stages. First, with iOS 27.3, Siri gets detailed control over Apple's own apps, such as rotating and cropping images in Photos or organizing Safari tabs. Second, likely in the spring, comes integration with third-party apps. Siri could then place a Grubhub order or book a table on OpenTable. Third, Apple wants Siri to handle ongoing tasks. Gurman explains that Dots, Muse and Grok can track goals, decide the next step and keep working without the user starting every step. He believes Siri could work this way as early as next year, 'not that it has much of a choice.'

This fits our thesis that the AI race is moving from model quality to infrastructure. Apple will not win on raw intelligence soon. What it controls is the operating system, permissions for over a billion devices, and the app integrations through which an agent actually acts. A less capable agent that can legally and technically tap every app on your phone may be more useful in practice than a smarter one stuck in a chat window.

There is a tension Apple itself has admitted. In its 10-K filing, the company said AI agents can carry out 'unauthorized' actions. Moving from cropping photos to placing orders and pursuing goals on its own sharply raises the cost of a mistake. A wrong food order is an annoyance. A wrong payment or booking made for you is a liability question.

Our reading: Apple's advantage should not be speed. Rivals already lead there. Its advantage could be making agents trustworthy: clear permissions, confirmation steps, logs you can review and actions you can reverse. If Apple builds that, it could set the de facto rules for consumer agents, as it has done with privacy and age checks. If it rushes to match Dots and Muse feature for feature under shareholder pressure, it gives up the one thing it does better than its rivals. In the long run, assistants that handle everyday chores give people time back. The question is who sets the guardrails while that happens.

🔗 Related on Zendoric

Sources & references