Apple sues OpenAI over engineer's alleged theft of confidential hardware files

Apple has filed legal action against OpenAI, seeking injunctions and damages over allegations that a former engineer weaponized an authentication flaw to pilfer confidential documents from Apple's servers weeks after leaving the company for the AI startup.

The Security Breach#

In early February 2026, Chang Liu—who spent eight years at Apple working on sensitive hardware initiatives—uncovered an authentication vulnerability on Apple's network while still retaining access to the company's systems via his work laptop. Rather than disclose the flaw, Liu allegedly exploited it to systematically extract dozens of confidential files related to unreleased products, circuit board designs, technical specifications, and internal engineering presentations.

According to Apple's court filing, Liu maintained this illicit access for several weeks while simultaneously developing hardware at OpenAI. In messages left on his Apple laptop, Liu appeared to joke about the vulnerability, writing "LOL" and commenting that he "found out I can access the [network storage], so funny" to a still-employed Apple colleague, Yu-Ting "Alyssa" Peng.

OpenAI's Alleged Role#

Apple claims the breach reflects a broader scheme involving OpenAI, asserting that the company and its executives deliberately enlisted departing Apple employees to "take an unlawful shortcut" toward manufacturing competitive AI-powered devices comparable to the iPhone. The lawsuit demands that OpenAI be blocked from using any Apple trade secrets and intellectual property in its hardware development.

Company's Response#

Apple states that it discovered the vulnerability through internal communications and promptly closed it. The company noted that while a small number of other users were exposed to the same authentication bug, server logs indicate they did not access or exfiltrate confidential information, distinguishing Liu's conduct from others affected by the flaw.