Apple seeks expedited discovery in OpenAI evidence destruction case

The evidence destruction allegation shifts the dispute from IP theft to procedural misconduct, potentially exposing OpenAI to sanctions that exceed the underlying infringement claim. This escalation suggests Apple's legal team believes the destruction itself is more damaging to prove than the original data transfer.

This case sits directly atop the employee theft allegations from earlier today (TechCrunch, 2026-09-01), where Apple accused a former employee of stealing data for OpenAI. The new filing adds a second legal front: if Apple can establish that OpenAI or its employees destroyed communications after learning of the investigation, courts may impose adverse inferences (treating destroyed evidence as proof of guilt) or sanctions that bypass the need to prove the underlying theft. This pattern mirrors how frontier labs are deploying legal playbooks to defend training data, as noted in that earlier coverage. The broader signal is that talent mobility disputes in AI are now fought on multiple fronts simultaneously: employment law, IP protection, and evidence preservation.

If Apple's expedited discovery motion is granted within 30 days and yields communications showing OpenAI employees coordinated the MacBook's late disclosure, expect OpenAI to either settle or face adverse inference rulings that assume the destroyed evidence was incriminating. Watch whether other AI labs preemptively adopt formal data retention policies in response to this precedent within Q4 2026.