As promised last month, Apple on Monday announced changes to Mac security controls designed to protect users’ data from the growing popularity of AI agents and their apps’ demand for “full disk access.”
“Before granting an app such extraordinary access, users should understand that it puts all of the data on their Mac at risk,” Apple wanred in a new blog post aimed at app developers.
Unlike AI chatbots that simply answer questions, agents run on a user’s device where they can take actions across the operating system, files, and apps on the user’s behalf.
While the tools can be useful, Apple has grown concerned about the increased risks to users’ data and privacy that these AI agents introduce. At the same time, the company has been benefiting from sales of Mac minis and Mac Studio computers that AI early adopters have been purchasing specifically to run their local AI models.
To balance these competing interests, Apple said in October that it would revise some of its security settings around Mac apps. The announcement followed two notable incidents involving AI agents. In one, a journalist claimed that Meta’s Muse app read their private messages without consent; a separate report from Wired cited a flaw in ChatGPT’s Mac app that could have allowed hackers to access sensitive data.




