Apple Vision Pro Vulnerability Allows Hackers to Flood Environments with 3D Objects

A recent security vulnerability in the Apple Vision Pro headset has left users startled as hackers were able to flood their environments with animated 3D objects, such as spiders and bugs, through a Safari exploit. The objects persisted even after exiting the browser, creating a uniquely unsettling experience for users.

Security researcher Ryan Pickren discovered the flaw (CVE-2024-27812) in February and disclosed it to Apple, who addressed the issue this month and awarded Pickren a bounty. The bug exposed the challenges in securing ‘spatial computing’ devices like the Vision Pro.

Apple had implemented strict privacy controls in the Vision Pro, including limiting apps to a default ‘Shared Space’ and requiring explicit user consent for immersive content. However, the AR Quick Look feature introduced in 2018 for iOS remained active in the visionOS without proper safeguards, allowing malicious websites to spawn unlimited 3D objects without user interaction.

Pickren explained that by adding specific anchor tags to webpages, hackers could fill a user’s room with crawling spiders and screeching bats without their knowledge. The exploit code is straightforward, and even closing Safari doesn’t remove the 3D objects, as they are handled by a separate application.

The researcher emphasized the need for Apple to reevaluate their threat model for the Vision Pro, as the bug’s impact goes beyond system crashes or reboots. This incident highlights the challenges in triaging and classifying vulnerabilities in emerging fields like Spatial Computing, raising concerns about security and privacy in technology.