TodaySaturday, November 21, 2026

Signal Blocks Screen Capture to Shield Chats From AI Assistants

The iOS 27 and supported-Android protections target AI access to readable chats, not a break in encryption, and leave notification and filesystem risks unresolved.
November 21, 2026
1 min read
A person holds a phone displaying the Signal logo against a blue Signal-logo background
The Signal logo displayed on a smartphone. Source: Reuters.

Signal is restricting screenshots on iPhones running iOS 27 and AI-triggered screen capture on supported Android devices to prevent assistants from reading private conversations. The messaging service announced the measures on November 20, while warning that blocking screenshots cannot stop every way an AI agent might access messages, according to Mashable and Signal’s announcement.

The protections work differently across the two mobile platforms. On iOS 27, Signal’s Block Siri AI setting is enabled by default, Mashable reports. On supported Android devices, Signal automatically enables Screen security when it detects an AI assistant trying to access the screen. Signal says differences among Android manufacturers prevent it from reliably implementing the same detection on every device. The announcement is therefore not a blanket screenshot ban covering all Android phones.

Users can disable the protections in Signal’s settings, but the company warns that doing so could expose both their own messages and those of their contacts to AI agents. The issue is separate from breaking end-to-end encryption, a protection also central to the rollout of encrypted RCS messaging between iPhone and Android. Messages must become readable on a recipient’s device, and an assistant with access to the screen or notifications may obtain that readable text without defeating the encryption used to deliver it. Signal says it cannot control what happens to data after an outside agent captures it, including whether it is processed locally or sent to a remote server.

The announcement follows reporting about Meta’s Muse assistant accessing Signal content. Mashable, citing Reuters, describes tests conducted with Signal and the AI Now Institute in which Muse obtained message text from Android notifications and used accessibility tools and screenshots to read chats on a Mac. Meta told Reuters that Muse can access other applications only through permissions users explicitly grant. Those examples involve device permissions and access to already-readable content, not evidence that Signal’s message encryption was cracked.

Signal President Meredith Whittaker is calling on Apple, Google and Microsoft to give developers documented ways to exclude their applications from AI access through screens, notifications, files and other operating-system interfaces. She also urged clearer disclosure of what assistants can read. Signal previously introduced default screenshot protection on Windows 11 to keep Microsoft Recall from capturing its chats. Separately, the service has faced regulatory scrutiny in India over its username feature, a different issue from AI access to message content.

The latest measures leave important gaps. Signal says screenshot blocking does not address potential access through files already stored on a device, and it cannot reliably block screenshots on macOS. For macOS 27 users, the company recommends adding Signal to Excluded Apps under System Settings and Siri to limit proactive access by Siri AI. The mobile changes reduce particular screen-capture risks; they do not establish that assistants can no longer obtain Signal content through other permissions or operating-system features.

Jennifer Hicks

Jennifer Hicks

Jennifer Hicks is a journalist at The Eastern Herald covering politics, geopolitics, international security, wars and conflicts, with a focus on major geopolitical developments and global security affairs.

Leave a Reply

Don't Miss