Signal closes one door to AI agents: screenshot blocking cannot protect everything

Chat privacy is not only about how messages travel: it also concerns who can read the screen once those messages are open.
On 20 November Signal announced new restrictions on conversation capture by agents built into mobile operating systems; The Verge covered the decision the same day.
This is yesterday’s update, not a feature announced today.

## What is being blocked
Signal describes blocking in-app screenshots on iOS 27 and taking action on Android versions where it can detect an AI assistant attempting to read the screen.
Its technical note says Android’s “Screen security” setting is enabled automatically when that access is detected, while acknowledging that not every device can be covered.
This is therefore not a uniform safeguard for every phone or operating-system version.
Users can disable the protection in the app’s settings, but Signal warns that doing so can expose their own and their contacts’ content to agents.

## The boundary extends beyond the screen
Signal itself calls the remedy incomplete: blocking screenshots does not stop an agent retrieving data already present in the device’s filesystem, for example.
The Verge highlights this distinction and the lack of robust tools for developers seeking to exclude their applications from AI access.
This is not an announcement that encryption has been broken: it concerns capturing content on a device after it has become visible.
Conflating those two layers would produce a conclusion the sources do not support.

## The same barrier is not available on Mac
Signal says it cannot reliably implement the same screenshot protection on macOS.
For macOS 27, it recommends adding Signal to Excluded Apps in Siri’s system settings to protect message content from proactive Siri AI access: a specific recommendation, not a general block against every agent.
We have not tested the control on a device and do not present the announcement as independent certification of invulnerability.

Signal’s request to operating-system providers goes beyond a single toggle: documented tools to opt applications out of AI data access and clearer technical information about how agents work.
The editorial question is who sets access boundaries when the operating system becomes an assistant: a private message also involves people who did not choose that assistant.

Report article

Back to WebSuggestion home