Apple and Google tighten data permissions to combat malware and AI privacy risks

Tech giants Apple and Google are addressing privacy concerns regarding how applications, particularly AI assistants and malware, access sensitive user data through broad system permissions.

On Android, hackers are increasingly exploiting Accessibility Services to gain control over devices. By tricking users into enabling these services, malware can monitor screen content, interact with user interface elements, and perform actions on the user's behalf. This technique is frequently used in banking trojans to execute overlay attacks, where fake login screens are placed over legitimate banking apps to steal credentials. In response, Google is introducing stricter limitations for apps requesting these services in Android 17 under Advanced Protection mode.

Meanwhile, Apple is planning to tighten controls over the ‘Full Disk Access’ permission in macOS. This follows reports that AI assistants, such as Meta’s Muse, could potentially access private messages, emails, and web history due to the broad scope of this permission. While users may intend to grant access to specific files, the permission can technically allow an application to read entire communication histories. Apple aims to implement more explicit confirmation steps to ensure users fully understand the risks before granting such extensive access.