Apple has announced plans to strengthen privacy controls around macOS’s “Full Disk Access” feature, citing growing security risks associated with increasingly capable AI agents.
The announcement comes amid renewed concerns about desktop-based AI applications gaining access to users’ private files, messages and other sensitive information.
Apple said the Full Disk Access setting, which was originally intended to allow applications such as backup tools to function properly, can give apps extensive access to data stored on a Mac.
The company said some developers are now using the permission in ways that could expose users’ information without them fully understanding the implications.
“Some developers are using Full Disk Access in ways that could put users at risk, exposing everything on their systems…without users’ full knowledge and understanding,” Apple said in a blog post directed at developers.
The company said it would introduce additional safeguards requiring users to take more explicit actions before granting an application Full Disk Access.
Apple said the move was necessary as AI agents become more autonomous and capable of interacting with files, messages and other content on users’ computers.
“As AI agents become increasingly capable and autonomous, the risks associated with this level of access will grow substantially,” the company said.
The announcement followed a report by Inc. columnist Jason Aten, who alleged that Meta’s Muse AI application was able to access the contents of his private messages despite his claim that he had not authorised such access. Meta disputed the allegation.
The controversy has intensified concerns about the security implications of desktop AI agents, which can perform tasks on behalf of users and, depending on the permissions granted, access personal information stored on their devices.
In the case of Muse, users can optionally grant the application Full Disk Access. According to Apple, the permission can allow an application to access files, emails, messages and browsing history.
Apple’s move also follows a Wired report about a vulnerability in OpenAI’s ChatGPT Mac application that reportedly could have exposed sensitive information.
The company said its new controls would be designed to ensure that users who genuinely want to give an application extensive access to their computers do so through “very explicit user action.”
Apple did not immediately respond to inquiries about when the new controls would take effect.
- Amaechi vows to back Tinubu if petrol price drops to N600 - October 3, 2026
- Apple tightens Mac privacy controls over growing AI agent risks - October 3, 2026
- US: Sanders moves to ban federal agencies from using Flock surveillance - October 3, 2026
