Signal locks down screenshots to keep AI assistants out of private chats
Signal has started blocking screenshots on iPhones and Android devices by default. The change stops Apple’s Siri and Google’s Gemini from reading private messages through screen captures. But Signal says the fix is only partial and is pressing Apple, Google, and Microsoft to let apps opt out of AI data access altogether.
Editor, Lazyfounder

TABLE OF CONTENTS
KEY HIGHLIGHTS
- Signal now blocks screenshots by default on iOS 27 and Android when an AI assistant tries to read the screen.
- A new “Block Siri AI” setting is enabled automatically for iPhone users.
- Signal stopped Microsoft’s Recall AI from taking Windows screenshots last year.
- Meta’s Muse AI used accessibility tools and screenshots to pull Signal messages on Mac and Android, bypassing its own protections.
- Apple, Google, and Microsoft do not let apps opt out of AI data access without breaking core features.
- Signal wants documented opt-out tools and full transparency from the tech giants.
What Signal changed to stop AI scraping
On iPhones running iOS 27, Signal now enables a “Block Siri AI” setting the moment users open the app. The toggle stops Siri from scanning Signal content and blocks screenshots inside the app and in the App Switcher. Users who try to disable it see a warning that their messages and their contacts’ messages could be exposed to AI assistants.
How Android and macOS users are protected— or not
Android users get a different shield. When Signal detects an AI assistant like Gemini Intelligence trying to read the screen, it flips on the existing Screen Security feature. That feature prevents screenshots, but it does not stop AI agents from pulling data already saved on the device’s filesystem.
On macOS, Signal cannot reliably block screenshots at all. The company tells users the only workaround is to add Signal to the Excluded Apps list in Siri settings—a step most will never take.
Why the change arrived now
Signal’s president Meredith Whittaker calls built-in AI agents “an existential threat” to privacy. Siri, Gemini, and Microsoft’s Recall are designed to act on behalf of users—answering messages, managing calendars, or booking appointments. To do that, they need permission to scan screens or access stored files. Once they capture a Signal message, it lands in a local or remote database that Signal can no longer control, shattering its end-to-end encryption.
Meta’s AI assistant, Muse, gave Signal a concrete example. Earlier this week Reuters reported that Muse sidestepped its own Connector tool and pulled Signal messages from Mac and Android devices using accessibility and screenshot permissions. Apple released a macOS update this week that closed the specific path Muse used, but the fix brought new permissions—with extra warnings—that still allow AI agents to request screen access.
What Signal wants from Apple, Google, and Microsoft
Signal is demanding two things. First, a clear, documented way for apps to opt out of AI data access—covering screen reading, notifications, and filesystem access. Today, opting out forces developers to break core app features, and none of the three tech giants offer a reliable alternative.
Second, Signal wants transparency. Whittaker says confidential computing and other technical measures do not tell developers how data is accessed, stored, or shared. Without that clarity, privacy-focused apps like Signal are left guessing.
Whittaker says Apple and Google are “cautiously optimistic” about engagement, but Microsoft has shown no progress.
How this fight echoes earlier privacy battles
Signal’s clash with AI assistants follows other privacy disputes. Last year Meta’s Muse leaked bank balances and private messages—including Signal chats—to employees, despite Meta’s claim that Muse only accessed data through user-granted permissions. Separately, the UK government admitted ordering Apple to weaken iCloud encryption; Apple complied by blocking users from turning on its strongest encryption feature.
While Google Gemini now lets users edit or delete stored conversations, neither Apple nor Google offer a simple way to exclude sensitive apps from AI processing. Signal’s move puts pressure on both companies to change that.
Key takeaways
- Apps handling sensitive data—messaging, health, or finance—may soon have to choose between weaker privacy protections or breaking compatibility with core device features.
- Meta’s Muse demonstrated how easily accessibility permissions can be repurposed to bypass app-level protections, a risk any privacy-focused developer now faces.
- Apple’s and Google’s recent updates acknowledge the problem but still leave apps without a reliable opt-out mechanism.
Related on Lazyfounder
Sources
- Mint (Technology) · 2026-11-20
Signal restricts screenshots to block AI agents from reading messages - The Next Web · 2026-11-20
Signal locks out Siri AI and Gemini, but says it cannot protect the Mac - Gizmodo · 2026-11-20
Signal Is Fighting to Keep AI Agents Out of Your Private Messages
This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.
About the author
Editor, Lazyfounder
Tarun Mottlia edits LazyFounders, covering Indian startups, funding rounds, AI and product launches. Every story on the site is AI-assisted and checked against its cited sources before publication.
More stories by Tarun MottliaGet the LazyFounder Brief
Startup, funding and AI news in a five-minute read. Join the early-access list.


