Sources
Join the community
Create your free Charmloop account — no credit card, no limits on browsing. Start making AI art in minutes.
Discuss this with
Pick a companion and get their take on this story

Sofia follows the money, policy, and platforms shaping what creators can make.
Create your free Charmloop account — no credit card, no limits on browsing. Start making AI art in minutes.
Pick a companion and get their take on this story
Signal, the encrypted messaging service, has updated its Android and iOS applications to block AI agents like Google Gemini and Apple's Siri from reading or analyzing user screens. This measure targets the increasing integration of agentic AI capabilities within operating systems that can interpret on-screen information, directly impacting user privacy.
Signal's president, Meredith Whittaker, stated that these agentic AI features are "undermining Signal's ability to independently guarantee privacy," according to The Verge. The core issue lies in the operating system-level access these AI agents now possess, allowing them to capture and interpret screen content without explicit user consent for each instance. For creators working with sensitive information or personal projects, this shift in OS-level AI access presents a new privacy frontier.
The privacy feature is enabled by default for new Signal installations, while existing users will be prompted to activate it. Users retain the option to disable the screen-blocking setting within the app's privacy menu if they prefer. This functionality extends beyond simple screenshots, specifically targeting the advanced contextual analysis performed by AI agents. For instance, Google's Gemini, especially with its recent "Computer Use" feature on macOS, is designed to understand and interact with applications and data across the operating system, a capability Google's Gemini app for macOS Gains 'Computer Use' Feature for Cross-Application Automation previously detailed.

Signal's Android privacy settings showing the screenshot blocking option.
Image: The Verge / The Verge AI
This move by Signal highlights a broader industry concern. As AI models become more integrated into operating systems, their ability to observe and process user activity expands. This could include sensitive data like private messages, creative workflows, or even proprietary design elements used in AI art generation. While convenient for users who want seamless AI assistance, it poses a direct challenge to applications built on strong privacy guarantees.
The decision by Signal sets a precedent for how other privacy-focused applications might respond to the proliferation of on-device AI. It forces a clear delineation between the utility offered by AI agents and the user's expectation of data isolation within specific applications. The tension between open access for AI agents and application-specific privacy controls will likely escalate as Apple, Google, and others continue to embed AI deeper into their platforms. Creators who use AI art tools often rely on various applications for inspiration, reference, and collaboration; the risk of an AI agent inadvertently accessing or analyzing creative works or sensitive prompts underscores the need for robust privacy controls. This also echoes earlier concerns about AI's potential for data leakage, which some platforms have addressed with features like Canva's AI 'off' switch for generative functions, as noted in Canva Adds AI 'Off' Switch for Image Generation and Magic Expand Features.
This development is a direct response to the industry's push for more pervasive AI. As on-device AI agents become increasingly sophisticated, developers of privacy-centric applications will face ongoing pressure to implement similar safeguards or risk undermining their core value proposition. The choice presented to users—convenient AI assistance versus segmented privacy—will shape the adoption and trust in next-generation operating systems.