Apple says it’s tightening macOS ‘Full Disk Access’ controls due to new risks from AI agents

Your apps may require more explicit permission to access your files and messages.

Apple announced it is implementing stricter controls for its macOS 'Full Disk Access' setting, a move prompted by concerns over new risks posed by AI agents. This decision follows reports of AI applications, such as Meta's Muse, potentially accessing user data like private messages without explicit consent, even when Full Disk Access was not fully granted. Apple stated that AI agents increase the risks associated with this high level of system access, which was originally intended for functions like backups. The company plans to require "very explicit user action" for apps seeking this level of permission, ensuring users fully understand the implications before granting it. Apple also cited a Wired report about a vulnerability in ChatGPT's Mac app that could have exposed sensitive data.

The enhanced security measures signal Apple's proactive stance in addressing the evolving threat landscape presented by increasingly sophisticated AI agents. By demanding explicit user consent for Full Disk Access, Apple aims to prevent inadvertent data exposure and reinforce user privacy in the face of autonomous AI functionalities. This development highlights a growing tension between the convenience of AI-powered applications and the fundamental need for robust data protection on personal devices.

This move by Apple will likely influence how other operating systems and application developers approach AI integration and data access permissions. Developers of AI agents will need to re-evaluate their data handling practices and consent mechanisms to comply with stricter guidelines, potentially leading to more transparent and user-centric design in future AI applications. Users, in turn, can expect a more granular control over their sensitive information when interacting with desktop-based AI tools.