Technology news and discussion on TalkFuse.
Signal Calls OS-Level AI an ‘Existential Threat’ and Blocks Screenshots to Fight It
Signal is adding protections to prevent smartphone AI assistants from reading and storing its users’ messages, saying screen-scanning features could expose private conversations beyond the app’s encryption. Its president, Meredith Whittaker, described operating-system AI integrations as an “existential threat” to Signal’s privacy mission and urged Apple, Google and Microsoft to give developers clearer controls over AI access.
Signal will block new agentic AI features in iOS 27 and automatically enable its screen security setting on Android when it detects an assistant trying to read the screen. Users can turn the protection off, but Signal warns this could expose their messages and contacts’ messages; the measure also cannot prevent access through a device’s filesystem and is not available on macOS. The company cited concerns about third-party agents using accessibility permissions, including a reported case in which Meta’s Muse accessed private iMessages.
In short
- Signal will block some AI assistants from reading messages on screen.
- Android protection turns on when an assistant tries to access Signal.
- Signal says the safeguard has limits and calls for clearer AI controls.
AI Gadgets Software Technology