Signal said in a blog post by its president, Meredith Whittaker, that it now restricts AI assistants built into phone operating systems from reading its screen. On iOS 27, the app blocks screenshots. On Android, it automatically turns on its "Screen security" setting when it detects an AI assistant reading the screen. Signal named Google's Gemini Intelligence and Apple's Onscreen Awareness with Siri AI. Users can switch the protection off in settings.
Signal says these assistants scan what is on screen, and most get app data through screenshots. The captured content can then sit in a local database or on a remote server that Signal cannot control. Blocking screenshots cuts that path, but not the filesystem route, where content already on the device could be read. On many Android makers' phones, Signal cannot reliably detect a screenshot attempt when screenshots are already off. It also did not want screenshots to simply stop working and confuse users. So it detects AI assistant access instead, and not every device is covered. It does not block Siri AI on macOS.
Signal's argument is that its own encryption cannot guarantee privacy once an operating system assistant captures the screen. Leaders could ask which apps hold regulated or confidential data, whether vendors can opt out of OS AI access, and what staff may do with such apps on their devices.