OpenAI Fights Apple Trade Secrets Suit by Exposing Apple’s Own Security Gaps

TL;DR

• OpenAI has filed new legal exhibits arguing that Apple’s trade secret claims are undermined by Apple’s own negligent security and offboarding practices, including evidence that an Apple manager accessed a former engineer’s iCloud account after departure.
• The counter-strategy shifts focus from whether OpenAI misappropriated secrets to whether Apple took “reasonable measures” to protect them—a legal requirement for any trade secret claim.
• If the court accepts OpenAI’s arguments, Apple could face a higher burden of proof, potentially weakening its injunction request and exposing broader corporate security failures.

The Backstory: A Lawsuit Built on a Single Departure

The legal war between Apple and OpenAI began when a senior Apple engineer, allegedly working on Apple’s confidential AI chip and battery design projects, resigned to join OpenAI’s competitive hardware division. Apple filed suit within weeks, claiming the engineer downloaded proprietary files to personal devices and that OpenAI encouraged the theft to fast-track its own silicon development. The initial complaint painted OpenAI as a predatory poacher, but the newest filings tell a very different story—one that places Apple’s own security apparatus under the microscope.

OpenAI’s New Exhibits: The iCloud Access Bombshell

The centerpiece of OpenAI’s latest legal response is a set of internal Apple documents and depositions that allegedly show a senior Apple manager used the departed engineer’s corporate iCloud credentials after the engineer’s termination. According to the exhibits, the manager accessed email archives, calendar entries, and shared design folders for weeks after the employee’s last day—ostensibly to “transition projects,” but OpenAI argues this proves Apple’s own offboarding process was chaotic and uncontrolled.

More damaging, the exhibits reportedly show that Apple’s IT department never revoked the engineer’s access to certain internal wikis and Slack channels until 72 hours after the lawsuit was filed. OpenAI’s lawyers argue that this means Apple cannot claim its secrets were “secret” at all—if a manager can casually log into a former employee’s account, then the company’s own conduct blurs the line between authorized access and theft.

Why This Matters: The “Reasonable Measures” Test

Under both California’s Uniform Trade Secrets Act and federal law (the Defend Trade Secrets Act), a plaintiff must prove the information was the subject of “reasonable efforts to maintain its secrecy.” This is not a strict liability standard, but it requires companies to show they actually enforced access controls, monitored departures, and audited data exfiltration.

OpenAI’s strategy is brutally simple: if Apple’s own manager accessed a departed engineer’s iCloud account without formal authorization, then Apple cannot credibly argue that the engineer’s later access (or any alleged download) was an extraordinary breach. Instead, OpenAI will frame the engineer’s actions as routine, sloppy, and consistent with Apple’s own internal culture of lax oversight. The exhibits also reportedly include Apple’s internal security training slides that explicitly warned against “password sharing” and “post-employment access,” which OpenAI will use to show that Apple knew the risks but failed to enforce its own policies.

The Counter-Strategy: Turning Plaintiff Into Defendant

OpenAI’s legal team is not just defending; they are on the offensive. The exhibits are designed to accomplish three goals:

1. Shift the burden of proof – By showing Apple’s own negligence, OpenAI forces the court to question whether the alleged “secrets” were ever protected. If Apple cannot prove reasonable measures, the entire lawsuit collapses on summary judgment.

2. Expose discovery abuse – OpenAI alleges that Apple’s initial discovery requests were overly broad, but Apple’s own records show they failed to preserve key server logs from the engineer’s departure period. This could lead to spoliation sanctions against Apple.

3. Undermine the injunction – Apple is seeking a preliminary injunction to bar the engineer from working on any AI or chip projects at OpenAI. OpenAI argues that if Apple’s security was this porous, there is no irreparable harm—the “secret” was already widely accessible within Apple’s own ecosystem.

What the Exhibits Reveal About Apple’s Internal Controls

Beyond the iCloud access, the filings reportedly detail a pattern of systemic failures:

No exit interview security checklist – The engineer’s exit interview was conducted by HR, not by a security officer, and no one verified that all corporate devices were wiped.

Shared admin passwords – The exhibits include internal chat logs where the engineer’s team used a common admin password for a design database, meaning at least six other employees could access the same files.

Personal device policy loopholes – Apple’s BYOD policy allowed employees to sync corporate email and files to personal iPhones and Macs, but the policy did not require remote wipe upon termination. The engineer’s personal laptop was never re-imaged until after the lawsuit was filed.

A manager’s “helpful” access – The most explosive exhibit is a series of calendar invites and emails showing the manager who accessed the iCloud account was also copied on the engineer’s personal resignation letter—meaning the manager knew the departure date but still used the credentials.

Industry Reaction: A Cautionary Tale for Big Tech

Legal experts not involved in the case say OpenAI’s move is a masterclass in defensive litigation. “Trade secret cases are won or lost on the plaintiff’s own security posture,” says one former federal prosecutor who specializes in IP. “Apple has always projected an image of fortress-like secrecy, but these exhibits suggest that internally, it’s the same chaos as any other tech company. If OpenAI can prove that Apple’s own manager accessed a departed employee’s account, that’s not just embarrassing—it’s fatal to the claim.”

Another analyst points out that this could have broader implications for how companies handle offboarding. “Every tech firm is now watching this case. If Apple loses, every future trade secret lawsuit will be met with a deep dive into the plaintiff’s own access logs. That’s a huge deterrent for companies that don’t have airtight security audits.”

What Happens Next

The court has scheduled a hearing for late August on Apple’s motion for a preliminary injunction. OpenAI has requested that the judge consider the new exhibits before ruling. Apple’s response is due within two weeks, and its lawyers will likely argue that even if a manager accessed the account, that does not excuse the engineer’s alleged bulk download of source code and proprietary schematics.

But the strategic damage is already done. The narrative has shifted from “OpenAI stole our secrets” to “Apple can’t keep its own secrets.” If the judge allows discovery to continue, OpenAI’s legal team will likely depose Apple’s security chiefs and demand a full audit of every offboarding since 2020. That could expose hundreds of similar incidents, turning a single trade secret suit into a corporate-wide security nightmare.

The Bigger Picture: Trade Secrets in the AI Era

This case is unfolding against the backdrop of an unprecedented talent war between Apple and OpenAI. Apple is reportedly building its own large language model and custom AI accelerators, while OpenAI is expanding into hardware with a former Apple chip lead. The stakes are enormous—billions in R&D, market timing, and proprietary architecture.

But this lawsuit is also a warning to the entire industry: if you sue a competitor for stealing secrets, be prepared to prove that your own house was in order. OpenAI’s exhibits are a reminder that in the digital age, “trade secret” is not a label you slap on a file—it is a discipline you practice every day, especially when employees walk out the door. Apple may still win on the merits, but its armor is now visibly dented, and the burden of proof has quietly shifted.

AndroGuider Team

Articles written by the AndroGuider team. We try to make them thorough and informational while being easy to read.