Apple Accuses OpenAI of Destroying Evidence in Trade Secrets Dispute

Evidence destruction in litigation can trigger sanctions ranging from adverse inferences to dismissal of claims.
Apple's allegation raises questions about what happens when parties fail to preserve materials relevant to their case.
Mark

What exactly is Apple saying OpenAI destroyed, and how did Apple find out about it?

Mimi

The reporting doesn't specify which documents or materials Apple believes were eliminated. We know it's tied to the trade secrets dispute, so likely emails, code, training data logs—things that would show who accessed what and when. Apple probably discovered it during discovery, when both sides exchange materials.

Luke

Right, and that's the gap we need to flag. The Bloomberg piece doesn't detail what evidence Apple claims is missing, or OpenAI's explanation. We're working from an accusation, not from a court finding.

Mark

Why does evidence destruction matter so much in a case like this?

Mimi

Because trade secrets cases live or die on documentation. You need to prove someone had access to confidential information, understood it was confidential, and then used it without permission. If the records that would show that access are gone, the other side's case gets much weaker.

Luke

Though we should note: Apple hasn't proven destruction happened. They've alleged it. OpenAI might have a legitimate explanation—system failures, retention policies that predate the dispute, something else entirely.

Mark

What happens if a court agrees with Apple?

Mimi

Sanctions. Could be an adverse inference, where the judge tells the jury to assume the destroyed evidence would have helped Apple. Could be dismissal of OpenAI's defenses. In serious cases, referral to prosecutors.

Luke

But again, we don't know the scope yet. Is Apple saying OpenAI destroyed everything, or specific materials? Is this negligence or intentional? The remedy depends entirely on those details, which aren't in this reporting.

Mark

Does this tell us anything about how AI companies handle their data?

Mimi

It suggests at least one company may not have robust systems for preserving materials once litigation becomes foreseeable. That's a real problem in an industry built on data.

Luke

Or it tells us Apple believes that. We don't have OpenAI's side, we don't have technical details, we don't have a court's preliminary assessment. This is one party's litigation position, not established fact.

  • Apple has accused OpenAI of actively destroying materials central to an ongoing trade secrets case — an allegation that, if substantiated, could expose OpenAI to sanctions far more damaging than the underlying lawsuit.
  • The charge introduces a legal crisis within a legal crisis: discovery, the process meant to surface truth, has itself become the site of alleged misconduct.
  • Courts treat evidence destruction with exceptional severity — intentional spoliation can trigger adverse inferences, claim dismissal, or even criminal referral, meaning OpenAI's litigation posture is now under as much scrutiny as its business conduct.
  • Apple's legal team appears to have uncovered signs of destruction during the discovery process itself, narrowing the window to preserve whatever evidence remains and accelerating the urgency of judicial intervention.
  • OpenAI has not publicly responded to the specific allegation, but the case is now poised to move through a gauntlet of discovery disputes before any trial — with the AI industry watching closely for what precedent emerges.

In the contested terrain where technology and law converge, Apple has accused OpenAI of destroying evidence in an active trade secrets dispute — an allegation that, if proven, would strike at one of civil litigation's most sacred obligations: the duty to preserve. The charge arrives at a moment when the artificial intelligence industry is already under scrutiny for how it acquires and handles the intellectual property of others, and it raises a question that courts may soon be forced to answer for an entire sector: what does good faith look like when the evidence itself is code, data, and the digital residue of machine learning?

Apple has alleged in court filings that OpenAI is destroying evidence in an active legal dispute over trade secrets, according to Bloomberg. The accusation targets one of civil litigation's most foundational obligations — the duty to preserve materials relevant to a case — and introduces a layer of potential misconduct that courts treat with particular gravity.

The underlying dispute appears to involve claims that confidential information was misappropriated, with the kinds of materials most relevant to such cases — emails, internal communications, code repositories, data logs — being precisely what Apple now says OpenAI is eliminating. Trade secrets cases turn heavily on access and intent, making documentary evidence not just useful but often decisive.

If a court finds that destruction occurred, the consequences depend on whether it was intentional, negligent, or inadvertent. Intentional spoliation invites the harshest remedies, including adverse inferences — where a court presumes the destroyed evidence would have supported Apple's claims — or outright dismissal of OpenAI's positions. The timing of Apple's accusation suggests its legal team may have uncovered signs of destruction through forensic analysis or the conspicuous absence of materials that should have existed.

OpenAI has not yet responded publicly to the specific allegation. The case is expected to move through a series of discovery disputes before any trial, with judges likely to issue preservation orders and potentially impose sanctions. How courts ultimately handle evidence obligations in AI litigation could establish lasting expectations for an industry that has increasingly found itself at the center of intellectual property conflict.

Apple has filed allegations that OpenAI is destroying evidence in an ongoing legal dispute centered on trade secrets, according to reporting from Bloomberg. The accusation strikes at the heart of litigation conduct—the obligation that parties preserve materials relevant to their case, a foundational principle of civil procedure that exists to prevent one side from erasing its own misconduct or weakening the other's ability to prove claims.

The specifics of what Apple contends OpenAI has destroyed, and the timeline of when such destruction allegedly occurred, remain central to evaluating the seriousness of the charge. Evidence destruction in litigation can trigger sanctions ranging from adverse inferences—where a court assumes the destroyed evidence would have supported the other party's case—to dismissal of claims or counterclaims, or even referral for criminal investigation if the conduct crosses into obstruction. Courts take these obligations with particular weight because discovery, the process by which parties exchange documents and information before trial, depends entirely on good faith compliance.

The trade secrets dispute between Apple and OpenAI appears to involve allegations that one company misappropriated confidential information belonging to the other. Trade secrets cases are inherently sensitive because they often turn on access: who had it, when they had it, what they did with it, and whether they took steps to keep it confidential. The documents and communications that would typically prove or disprove such claims—emails, meeting notes, code repositories, training data logs—are precisely the materials that Apple now says OpenAI is eliminating.

Apple's allegation introduces a secondary legal question layered atop the underlying trade secrets claim. If a court finds that OpenAI has indeed destroyed relevant evidence, it must then determine whether the destruction was intentional, negligent, or inadvertent—a distinction that carries enormous weight in how courts respond. Intentional destruction, sometimes called "spoliation," invites the harshest remedies. Negligent destruction might result in lesser sanctions. Inadvertent destruction, if the party can show it took reasonable steps to preserve materials, may result in no sanction at all.

The timing of Apple's accusation is notable. It suggests that Apple's legal team discovered evidence of destruction during the discovery process itself—perhaps through forensic analysis of OpenAI's systems, through testimony, or through the absence of materials that should have existed. This kind of discovery-within-discovery often prompts immediate legal action because the window to preserve remaining evidence narrows as time passes.

OpenAI has not yet publicly responded to the specific allegation, though the company has faced other legal challenges in recent years related to training data and intellectual property. The artificial intelligence industry more broadly has become a flashpoint for disputes over whether companies have properly licensed or obtained consent for the data used to train their systems. These cases often hinge on documentary evidence—licensing agreements, communications about data sourcing, internal policies about what material could be used.

The case will likely proceed through discovery disputes before any trial, with judges potentially issuing orders requiring both parties to preserve all relevant materials going forward and possibly imposing sanctions if destruction is found to have occurred. How courts handle evidence preservation in AI-related litigation may establish expectations for the entire sector, particularly as companies race to develop and deploy large language models and other systems that require vast amounts of training data. The outcome could shape whether companies face heightened obligations to maintain detailed records of their data sourcing and model development processes.

Contact Us FAQ