Signal is expanding its screenshot protections on mobile as it warns that operating system AI agents could reach information shown on a user’s screen. The company says the change is meant to limit exposure of private conversations, though it cannot prevent every possible route by which an AI agent might access app data.
Mobile screenshot limits are expanding
Signal announced Friday that it is adding screenshot restrictions in Android and iOS. The secure messaging app had already blocked screenshots by default in its Windows app.
The change responds to concerns about AI agents built into operating systems. Signal President Meredith Whittaker described those systems as an “exis tential threat” and said their ability to scan screen contents undermines Signal’s ability to guarantee privacy independently.
In practical terms, a conversation may be protected within Signal while information displayed on the screen is also available to the operating system. Signal’s concern is that an AI agent could capture that information and pass it to a database or server used by the operating system’s AI features.
Why screen access matters for private conversations
Signal’s privacy concern extends to everyone in a conversation, not just the person whose device is being scanned. A message shown on one participant’s screen can contain sensitive information shared by another participant.
Whittaker said that once data has been captured by an operating system AI agent, it is outside Signal’s control. That creates a gap between the privacy protections Signal can apply inside its own app and the access an operating system may have to what appears on a device’s display.
That distinction helps explain why screenshot controls matter to the company. Limiting screenshots can reduce one way that screen content is copied, but Signal’s warning is broader: AI agents that inspect screens may expose data beyond the app’s safeguards.
Users can turn off the restriction
The setting is optional. Users can disable it if they choose, but Whittaker said that doing so would “break our privacy guarantees.” The choice therefore affects the protection Signal can offer when sensitive information is visible on screen.
For users, the setting is a tradeoff between allowing screenshots and keeping this particular barrier in place. Signal’s position is that its privacy promise depends in part on restricting access to information shown inside the app, even though users retain control over the setting.
A safeguard with clear limits
Signal acknowledges that screenshot restrictions are an “incomplete” response. They do not address other, less visible ways that AI agents could potentially access application data.
That caveat matters because the new control addresses a specific risk rather than resolving the larger question of how operating system AI features interact with private apps. Signal can limit screenshots within its app, but it says it cannot control what happens after an AI agent captures information through another route.
The move extends a protection already used on Windows to Android and iOS. It also puts the responsibility for the setting partly in users’ hands, while highlighting a privacy boundary that Signal says operating system AI can cross.