Apple Sues OpenAI Over Stolen Hardware Secrets, and Some OpenAI Staff Are Fighting Back Against Their Own Boss

Key points

• Apple filed a lawsuit against OpenAI last Friday, accusing the company of taking unreleased iPhone parts, prototypes, and confidential project documents.
• OpenAI hired more than 400 former Apple employees, according to the lawsuit.
• OpenAI paid $6.5 billion in 2024 to acquire IO Products, a hardware startup co-founded by former Apple design chief Jony Ive.
• New York became the first US state to pass a statewide moratorium, a temporary ban, on new data centres, the large buildings full of computers that power AI services.
• A group of OpenAI employees started a Super PAC, a type of political fundraising group, to push for stricter safety rules on AI, putting them at odds with their own employer.

Apple is not a company that forgives quietly. Last Friday it sued OpenAI, claiming that former Apple staff carried confidential hardware secrets out the door and straight into OpenAI's new hardware division. The alleged haul includes unreleased iPhone parts, product prototypes, and documents about projects Apple had not yet announced publicly.

Central to the case is Tang Tan, OpenAI's chief hardware officer, who spent 24 years at Apple before jumping ship. The lawsuit claims Tan encouraged departing Apple staff to bring proprietary information with them. OpenAI declined to comment publicly before publication.

Apple has played this card before. As WIRED reported this week, longtime Apple executive Tony Fadell says Steve Jobs once threatened to sue him after he founded Nest, the smart-thermostat company, and hired hundreds of Apple employees. The lesson Fadell drew: it is Apple's job to keep talent, not just acquire it.

But Zoë Schiffer, a contributing editor at WIRED, frames the lawsuit as strategic rather than purely legal. Apple's deeper goal, the argument goes, is to slow OpenAI's hardware ambitions before they become a threat to the iPhone, which Apple is betting will remain the central device of the AI era.

OpenAI's planned device is reported to resemble a smart speaker with motorised parts. The idea is a voice-first gadget you talk to rather than stare at, handling tasks through an AI agent, software that completes multi-step jobs on your behalf. Whether that vision can actually beat a phone people already carry everywhere is a separate question.

Should ordinary users care about any of this?

Not immediately, but the background matters. If Apple wins in court and slows OpenAI's hardware push, the AI gadgets competing with your iPhone shrink in number. If OpenAI presses ahead, users could eventually face a genuine choice between ecosystems, not just apps.

Meanwhile, inside OpenAI itself, a group of employees launched a Super PAC this week to advocate for tougher AI safety rules, placing them publicly against company leadership. A Super PAC is a political fundraising group that can spend unlimited money on advocacy without coordinating directly with a candidate or party.

On the regulatory front, New York Governor Kathy Hochul signed the first statewide moratorium on new data centres. A data centre is a warehouse-sized building packed with the specialised computers that run AI models and store cloud data. The ban pauses new construction while the state studies the electricity and environmental costs. President Trump criticised the move this week, and several other states are watching to see whether New York's approach holds.

Separately, federal records obtained under freedom-of-information requests show that members of DOGE, the informal advisory group officially called the Department of Government Efficiency, used AI tools to help shape housing policy at the US Department of Housing and Urban Development. The government has so far refused to release details of how those tools were used.