Apple to tighten Mac privacy controls after complaints over Meta’s AI agent

• Courtesy of Apple

Apple is preparing stricter privacy controls for Mac computers following complaints that Meta’s artificial intelligence (AI) agent, Muse, accessed private messages without users’ knowledge.

The technology giant announced on Friday, Oct. 2, that it plans to update macOS to make it clearer when AI-powered applications request extensive access to information stored on a user’s computer.

“We give developers powerful APIs to build incredible capabilities into their apps for Apple products, backed by a set of controls designed to protect users’ private data,” Apple said.

“Full Disk Access largely sidesteps these controls in order to allow backup apps to function properly on the Mac. Some developers are using Full Disk Access in ways that could put users at risk, exposing everything on their systems—including files, mail, messages, and even browsing history—without users’ full knowledge and understanding. For communication apps, this can also compromise the privacy of the people users are communicating with,” It added.

The move follows concerns over Meta’s Muse, an AI agent designed to perform complex tasks on behalf of users, including canceling unused subscriptions and negotiating lower prices.

According to Reuters, some users alleged that Muse accessed messages they believed were private, raising questions about how much personal information AI assistants should be allowed to access.

Apple moves to strengthen data protection

Unlike iPhones and iPads, which use a security mechanism called sandboxing to prevent applications from accessing one another’s data by default, Mac computers offer broader permissions through a feature called Full Disk Access.

The feature allows authorized applications, such as cloud backup services, to access information stored across a computer.

However, Apple acknowledged that some developers have been using the feature in ways that could expose users to privacy risks.

The company plans to introduce additional safeguards requiring users to take explicit action before granting applications extensive access to their data.

Apple warned that increasingly capable and autonomous AI agents could make such permissions more dangerous, particularly when applications can independently retrieve and process sensitive information.

Meta disputes unauthorized access claims

The controversy gained attention after Inc. magazine technology columnist Jason Aten accused Muse of accessing private messages on his Mac, despite his claim that he had not enabled Full Disk Access.

Meta disputed the allegation.

Company spokesperson Andy Stone said Muse can access Apple’s Messages application only when users explicitly enable both Full Disk Access and the Messages connector.

Stone also maintained that users can revoke those permissions at any time.

The conflicting accounts have drawn attention to how AI agents handle operating system permissions and whether users fully understand the information these applications can access.

Growing privacy concerns over AI agents

Apple’s planned changes come as AI developers increasingly introduce agents capable of performing tasks across multiple applications and services with limited human intervention.

Unlike conventional chatbots that primarily respond to questions, AI agents can interact with applications, retrieve information and carry out actions on behalf of users.

These capabilities can require access to sensitive information, making permission controls increasingly important.

For Mac users, Apple’s upcoming safeguards are intended to provide greater transparency and control over which applications can access their personal data.

Apple has not announced a specific release date for the additional controls.

Source: Reuters