End-to-end encryption protects messages in transit and at rest. Once those messages render on screen, however, an AI assistant embedded in the operating system can read them just as readily as any secretly tracking users would.
What Signal Changed
A new default setting blocks screenshots and App Switcher previews to keep Siri out of your conversations.
Signal‘s latest iPhone update enables a “Block Siri AI” privacy setting on by default, according to 9to5Mac. The setting prevents screenshots inside the app and stops chat previews from appearing in the iPhone App Switcher. Attempt a screenshot anyway and Signal surfaces a message: “Screenshot prevented by ‘Block Siri AI’ setting.”
Why Screenshots Are the Target
Apple does not currently give developers precise tools to block AI agents, so Signal blocked screenshot and screen-preview access as a workaround.
The core problem is architectural. Signal told Reuters that Apple does not currently provide iOS developers with more granular controls to prevent Siri and third-party AI agents from accessing app content. Blocking screenshots and screen previews is a workaround, not a purpose-built solution. It is the most direct instrument Signal has available, given the absence of finer operating-system controls. For more on how Apple’s ecosystem can leave data files exposed, the iCloud case offers a concrete example.
The Trade-Off You Are Making
Turning the protection off is straightforward, but Signal warns the choice exposes your contacts as well as yourself.
Users who need screenshots can disable the Block Siri AI toggle inside Signal’s privacy settings. Signal cautions, according to The Verge, that doing so could expose not just your messages but those of your contacts to AI agents operating on your device. The toggle remains reversible, so the decision rests with each user , for more ways to stay safe on mobile, it helps to evaluate each setting carefully.
What This Protection Does Not Cover
Screenshot blocking addresses one access route; it does not close every path an AI agent might take.
This update is a mitigation, not a guarantee. According to The Verge, AI agents may still be able to reach content already stored on the device’s filesystem, meaning the protection does not eliminate every possible access route to Signal-related data. Signal’s own framing of the update acknowledges these limits rather than presenting the change as a complete defense.
Android Handles It Differently
Android devices receive a related protection, though the trigger differs from the blanket iOS default.
On Android, Signal uses a related screen-security mechanism that activates when it detects an AI assistant attempting to capture or inspect the display, according to Reuters. The underlying goal matches the iOS approach; the activation method does not. Because Android distributions vary, the behavior may differ depending on the device and operating-system version.
The Bigger Picture
AI assistants are expanding into the screen layer, where encryption offers no protection.
Privacy tools designed around transmission and storage are confronting a problem they were not built to solve. AI assistants embedded in operating systems now operate close to the display layer, where they can read content that encryption protected at every earlier point in its journey. The rise of the surveillance app threat landscape makes Signal’s response all the more significant. Signal’s update is an imperfect but concrete response to that shift, and the tension between OS-level AI access and app-level privacy controls is unlikely to resolve quickly.





























