Apple reveals new Mac security controls as AI agents raise privacy risks
Summary
- Apple announced changes to Mac security controls to protect user data from AI agents demanding full disk access.
- AI agents run locally on devices and can take actions across the operating system, files, & apps on behalf of users.
- The updates follow incidents involving Meta's Muse app reading private messages without consent & a flaw in ChatGPT's Mac app that could allow hacker access.
- Device Control & Data Access settings are split into System & App Protection Override & Accessibility to clarify security risks.
- A single method for granting permissions now requires selecting the requested app from a list.
- New guidance encourages users to grant access only to specific apps when enabling System & App Protection Override or Full Disk Access.
- Users will gain increased visibility into the range of data apps access or control when full permissions are granted.
- Users can configure apps to request permission every time access is needed for more granular control.
- A new transparency indicator appears when an app controls another app through the System & App Protection Override option.
- The changes arrive with macOS 27.3 & provide users with more nuanced options without preventing the installation of AI agents.
- Apple advises developers to adjust onboarding processes & handle limited folder or app access gracefully.
- Developers requesting Full Disk Access for disk backup or assistive technologies can request entitlements to preserve existing functionality.
Following an October announcement, Apple has detailed new permission controls, warnings, and transparency features coming to macOS 27.3 to address risks posed by AI agents and other apps.
- Keywords
- Apps,Privacy,Security,AI,ai agent,Apple,mac