Apple has taken two steps at once on AI agents this week. On 3 November 2026 it opened its rebuilt Siri AI to every app developer in a new preview, and in its annual report filed the same day it warned investors that AI agents could act without authorisation and expose the company to lawsuits. Together they show how a big platform is trying to move quickly on agents while flagging the risks.
What developers can do now
Apple released the 27.3 betas of iOS, iPadOS, macOS and visionOS with a Siri AI Developer Preview open to anyone in the Apple Developer Program. Using Apple's App Intents framework, any app can now expose its own content and custom actions to Siri AI, rather than only actions that fit Apple's predefined categories. Apps can also chain several actions into one request, and developers can teach Siri multi-step "skills" using simple Markdown files, according to coverage of Apple's documentation. Apple says the capabilities will reach customers in a few months, and App Store submissions for apps that use them open in spring 2027. Apple has not yet published the terms and usage policies that will govern what developers can build, and it has not given a price for access.
What users get and where
Siri AI is designed to use personal context such as emails, photos and messages, to understand what is on the screen, and to carry out tasks inside third-party apps. Apple has named WhatsApp and Microsoft Outlook among examples. It needs an Apple Intelligence-capable device. It currently supports English, French, Japanese, Korean, Portuguese and Spanish, with Traditional Chinese, Norwegian, Turkish and Vietnamese coming soon. It is not available on iPhone, iPad or Apple Watch in the EU and not available to users under 13.
The warning in Apple's filing
In its annual filing with the US Securities and Exchange Commission, Apple said AI agents could independently execute or change software code, send messages and complete transactions, and that in doing so they could reach systems inside or outside the company without the necessary permissions. It said such agents may behave in unintended or unpredictable ways, including taking actions that were not directed or authorised, and that its ability to monitor and control agents built by third parties could be limited. It warned this could lead to security, privacy and legal risks, including lawsuits and reputational damage, as reported by Anadolu Agency. On the company's earnings call, Apple executive John Ternus said Apple wants to make sure the technology is rolled out in a way that protects security and privacy, and stressed that users should understand which information and systems they are allowing agents to access. He said Siri could help users find personal information on their devices while keeping it private from Apple.
Why it matters
Apple is the latest company to say, in a formal legal document, that autonomous AI agents are a risk it cannot fully control. That matters because the iPhone is one of the world's largest app platforms, and apps that can be driven by an AI assistant create new ways for mistakes or abuse to spread, such as an assistant sending a message or making a purchase that the user never intended. The caution in the filing sits alongside the push in the developer preview, and the real test will be the rules Apple publishes before spring.
For app builders, including developers across Africa who build for iPhone users, the near-term step is to try the preview, think about which actions are safe to expose, and wait for Apple's policies before planning a release.
Sources: Apple developer announcement (3 November 2026), via iClarified, PPC Land, eWeek and DEV Community; Anadolu Agency; Crypto Briefing; Basic Tutorials.
