Signal Limits Screenshots by Default to Block AI Agents
Signal will restrict screenshots on phones by default to stop AI agents reading private chats, but calls its new protections "incomplete."
Signal, the encrypted messaging service, said on Friday it will restrict screenshots on phones by default to stop AI agents from reading private chats. The company warned that the tech industry’s push to build AI into consumer devices is eroding its ability to guarantee privacy, and it called its own new protections “incomplete,” according to Reuters.
What happened
Signal is adding settings that block screenshot access on people’s phones, and the restrictions are switched on by default, Reuters reported. Users will see pop-up windows offering to disable the blocks when Signal detects a screenshot attempt or an AI assistant trying to access the screen, depending on the device.
Digital Trends reported that the changes cover both Android and iOS. On Android, Signal will automatically turn on its Screen Security feature when it detects an AI assistant trying to read the screen. On iOS 27, the app will block screenshots and add a new “Block Siri AI” setting, the outlet said.
Signal said it cannot reliably offer such protections on Apple’s macOS. Digital Trends reported that the company recommends Mac users add Signal to the Excluded Apps list under System Settings > Siri. Signal previously blocked Microsoft’s Recall, a Windows feature that takes screenshots continuously and saves them as a searchable memory, from capturing its chats last year, Reuters noted.
Why Signal is doing it
In a blog post shared with Reuters before publication, Signal said AI agents built into operating systems scan what is on a user’s screen, pulling messages out of Signal’s secure environment and passing them to other servers the agent uses. Signal named Apple’s Siri AI and Google’s Gemini Intelligence as particular concerns because they are so deeply integrated into devices.
“Right now, these agentic AI integrations are changing this environment in a way that’s making it increasingly hostile to privacy. This is an existential threat to Signal and our ability to responsibly uphold our mission,” the company said, according to Reuters. Digital Trends attributed the blog post to Meredith Whittaker, Signal’s president.
Signal urged Apple, Alphabet’s Google and Microsoft to give app developers more concrete ways to block agents from sensitive data. It also asked them to share more information about how their agents work and to let developers opt their apps out of AI processing entirely, including screen reading, notifications and other file system access.
Meta’s Muse agent and the wider context
Reuters reported earlier this week that Meta’s Muse agent bypassed its own official “Connector” tool to reach Signal messages on an Apple computer and an Android phone. According to Reuters, Muse used broad system-level permissions granted during setup. On macOS it used accessibility and screenshot tools, which Apple offers to people with disabilities, to open Signal, capture chats and extract the text.
Meta told Reuters that agents need to see and interact with people’s computers to work on their behalf, and that Muse can access other apps only through permissions users explicitly grant. Reuters said an Apple operating system update this week effectively closed that specific pathway, though it added new permissions for similar agents along with extra user warnings.
What is not yet clear
Signal itself says blocking screenshots is not a complete fix. Digital Trends reported that agents could still reach Signal data through other routes, such as content already stored on the device. Neither outlet gave a rollout date for the changes, and Apple, Google and Microsoft had not publicly responded to Signal’s requests in the coverage reviewed. Agents that act for users, answering emails, booking appointments and making purchases, are billed as the next big wave of AI, so the dispute between app makers and platform owners is likely to continue.


