OpenAI Argues Apple’s Security Gaps Undercut Trade Secrets Claim

OpenAI is urging a court to throw out Apple’s trade secrets lawsuit, arguing that Apple’s own security and offboarding failures weaken the case that any protected information was actually stolen. The dispute matters because it pits two of Silicon Valley’s most influential companies against each other at a moment when AI hardware, talent, and proprietary product development are becoming increasingly valuable.

In newly filed court documents, OpenAI says Apple has not clearly identified the specific trade secrets it believes were taken and instead relies on broad descriptions of product-development know-how. The company also points to records it says show Apple workers kept access to personal iCloud accounts and that an Apple manager continued using a former engineer’s account after he had left the company.

What is the lawsuit about?

The case began when Apple sued OpenAI in July, accusing the AI company of participating in a scheme to obtain confidential hardware information from former Apple engineers. Apple’s complaint centers on the idea that former employees carried sensitive knowledge about product design, testing, suppliers and distribution into a rival company at the heart of the generative AI race.

Apple has argued that the alleged conduct was serious enough to warrant accelerated discovery, and this week it asked the court to move faster, saying its internal review suggests additional former employees may have been involved in, or at least aware of, the alleged misuse of confidential material.

OpenAI’s response does not simply deny wrongdoing. Instead, it tries to shift the legal framing. The company is asking the court to focus less on whether former Apple employees later joined OpenAI and more on whether Apple actually protected the information it now describes as trade secrets.

Why OpenAI is targeting Apple’s security practices

OpenAI’s central argument is that Apple cannot convincingly claim trade secret protection if it did not consistently secure the information or properly cut off access when people left the company. In legal terms, the argument is designed to undermine one of the basic requirements of a trade secret claim: that the information was subject to reasonable efforts to keep it secret.

According to OpenAI, Apple allowed employees to use personal iCloud accounts for work-related tasks and failed to fully revoke access after they departed. The filing says those practices created confusion over who could reach what data, making it harder for Apple to argue that it treated the material as tightly controlled and protected.

OpenAI also submitted text messages it says show a more complicated relationship than Apple’s complaint suggests. In those messages, an Apple manager reportedly remained logged into the personal iCloud account of former Apple engineer Chang Liu after he had left the company in order to move files, and later reached out for technical help with Apple-related work questions.

That detail matters because it supports OpenAI’s broader claim that some exchanges may have looked less like theft and more like lingering professional contact between former colleagues. OpenAI is attempting to show that the flow of information was messy, informal and perhaps even tolerated inside Apple, rather than the product of a clear, covert scheme.

How strong is OpenAI’s defense?

OpenAI’s defense is strongest as a challenge to Apple’s legal theory, not as a final answer to whether any improper disclosure occurred. The filing is meant to weaken the premise that the information at issue qualifies as a trade secret in the first place. If Apple cannot show strong protection measures and specific, identifiable secrets, its case becomes harder to sustain.

That does not necessarily end the matter. Courts can still find that trade secrets existed even when a company’s internal controls were imperfect. But OpenAI is betting that Apple’s alleged security lapses, employee-access problems and loose offboarding process will create enough doubt to narrow or derail the lawsuit.

OpenAI argues that it has no interest in Apple’s trade secrets and no need for them, saying the company is focused on building a different kind of product and wants talented engineers who choose to join it voluntarily.

The company also says Apple is trying to turn a talent dispute into a trade secrets case. In OpenAI’s telling, this is not really about stolen information at all. It is about Apple losing employees to a fast-growing AI company that is aggressively hiring people with hardware, product and systems experience.

What exactly did OpenAI say in court?

OpenAI’s filing repeatedly argues that Apple’s complaint is vague and overbroad. Rather than naming concrete secrets, OpenAI says Apple describes the allegedly stolen material in broad categories tied to ordinary product development.

OpenAI characterizes those categories as things like component manufacturing, product testing, vendor and supplier relationships, and distribution channels. In other words, the company is saying Apple has not yet pointed to a narrowly defined set of protected facts so much as a general collection of business knowledge that many senior engineers may naturally accumulate over time.

The distinction is important because trade secret law does not protect every piece of confidential-sounding information a company may possess. It protects information that is secret, derives economic value from being secret, and is subject to reasonable steps to preserve its secrecy. OpenAI is trying to show Apple has not met that standard.

OpenAI also claims Apple is using the litigation to slow down a rival that is building AI-powered hardware and systems of its own. The implication is that Apple wants to disrupt OpenAI’s momentum in a market where both companies may eventually compete more directly.

Why does the former-employee issue matter so much?

The former-employee issue is central because trade secret disputes involving high-skilled talent often hinge on what departing workers knew, what they took with them, and how their former employer handled access during and after departure. When employees leave for a rival, the line between expertise, memory and protected confidential information can get blurry fast.

Apple says some former workers may have participated in or observed the alleged theft of trade secrets, which is why it wants discovery to move quickly. OpenAI, by contrast, says the real story is that employees changed jobs and continued to communicate with former colleagues, which is not unusual in technology circles.

That clash reflects a larger reality in Silicon Valley: companies increasingly compete not just for products and customers, but for talent with deep knowledge of supply chains, manufacturing and AI systems. In a world where product cycles are compressed and hardware is increasingly intertwined with software, employee movement can trigger suspicion even when no physical documents are passed around.

How Apple’s internal controls became part of the fight

Apple’s internal controls are now part of the case because OpenAI says they are relevant to whether the information Apple complains about was truly safeguarded. The company points to the use of personal iCloud accounts and lingering access after employment ended as signs that Apple did not handle its own information with the level of rigor it now says was required.

That line of attack is not unusual in trade secret litigation. Defendants often argue that a plaintiff’s own security practices were too loose to support a claim that the material remained secret. The logic is simple: if access was broad, unmanaged or poorly documented, the secret may be less legally defensible.

Apple’s counterargument will likely be that imperfect security does not erase confidentiality, especially if the company had internal policies and expected employees to treat the information as restricted. The court will ultimately need to decide whether Apple’s controls were enough under trade secret law.

What does OpenAI say about its hiring practices?

OpenAI says it is simply hiring skilled people who want to work on new technology, and that Apple is attempting to recast ordinary recruitment as misconduct. The company insists it has no use for Apple’s proprietary information and no desire to incorporate it into its own systems or products.

That argument is aimed at separating talent mobility from secret theft. Companies in fast-moving sectors often hire rivals’ employees precisely because they value the expertise those people bring. The question is whether that expertise includes protected know-how or just general industry experience.

OpenAI’s filing suggests that some former Apple employees were drawn by the opportunity to work on a more ambitious or more innovative AI-related mission. The company’s position is that the movement of people should not automatically be treated as the movement of stolen secrets.

OpenAI says Apple should not use litigation to compensate for its weaker ability to keep talent and to integrate AI into its products.

That is a pointed allegation, and it shows this is about more than one legal complaint. It is also a reputational contest over which company is better positioned for the next era of consumer technology.

Why this dispute matters for the AI and hardware race

The case matters because it highlights one of the most consequential overlaps in tech right now: the convergence of AI software, consumer hardware and the talent needed to build both. OpenAI has already become a dominant name in software-driven AI, while Apple remains one of the most powerful product companies in the world, with a deep history in device design and manufacturing.

If Apple can show that confidential hardware information flowed to a rival, it could create pressure on companies trying to recruit experienced engineers from top device makers. If OpenAI prevails, it could reinforce the idea that companies must do a much better job securing their information before accusing departing employees and their new employers of theft.

The dispute may also influence how companies manage offboarding in the AI era. The more valuable the knowledge held by engineers, the more likely firms are to tighten access controls, audit personal account use and document exactly what leaves with a departing employee.

Possible wider effects

• Companies may review how they handle personal cloud accounts used for work.
• Employers could tighten access revocation when employees resign or are terminated.
• Trade secret complaints may need more precise descriptions of the alleged stolen material.
• AI hardware startups may face greater scrutiny when hiring from established device makers.

What happens next in court?

The next phase will likely focus on whether Apple can keep the case moving and whether OpenAI’s motion to dismiss persuades the judge that the complaint is too vague or legally weak. If the court allows the suit to proceed, discovery could become a major battleground over employee communications, access logs, offboarding procedures and internal records.

Discovery is often where trade secret cases either gain traction or begin to fall apart. The parties can test who had access, what was stored where, how it was shared and whether the plaintiff truly treated the information as secret. That is why Apple’s request for accelerated discovery and OpenAI’s pushback are so important.

Even if the case does not result in a public trial, the filings already offer a glimpse into how legal disputes over AI competition are likely to unfold. They are increasingly less about a single document or a single email and more about the broader systems companies use to manage people, data and control.

How the legal arguments compare

OpenAI and Apple are making fundamentally different claims about the same set of facts. Apple says there was a scheme to misappropriate confidential information. OpenAI says there was no protected secret, no meaningful concealment and no reason to assume the information was improperly stolen rather than loosely handled during employee transitions.

The table below summarizes the core positions:

What to watch next

The biggest question now is whether OpenAI’s motion to dismiss persuades the court that Apple’s complaint lacks enough detail and legal grounding to continue in its current form. If not, the case could move into a lengthy discovery process that may expose more about how both companies manage sensitive information and talent.

Another question is whether Apple’s internal review, which it says points to additional former employees, will strengthen the case or complicate it further. More names and more evidence could support Apple’s theory, but they could also widen the dispute and create new avenues for OpenAI to challenge the quality of the company’s controls.

For now, the fight is as much about narrative as evidence. Apple wants to frame the matter as theft of valuable proprietary know-how. OpenAI wants the court to see an employer with weak security practices, broad accusations and an attempt to turn competition for talent into a legal weapon.

Either way, the case underscores how fiercely major tech companies are defending their position in the AI era — not just in products and models, but in the people and processes that power them.

Timeline of the dispute

1. July 2026: Apple files suit against OpenAI over alleged trade secret misappropriation tied to former engineers.
2. Early August 2026: Apple asks for quicker discovery, saying its internal investigation suggests more former employees may be involved.
3. August 6, 2026: OpenAI files its motion to dismiss, attacking Apple’s security practices and the specificity of the complaint.
4. Next steps: A judge will decide whether the case proceeds, narrows or gets dismissed.

Bottom line

OpenAI is not merely denying Apple’s accusations; it is trying to dismantle the legal foundation of the lawsuit by arguing that Apple failed to protect the information it now calls secret. The result is a high-profile test of how trade secret law applies when elite tech talent changes employers in one of the most competitive moments in the history of artificial intelligence.