Fast And Loose
Meta and Apple filed lawsuits days apart accusing the AI industry’s largest companies of ignoring employment and intellectual property law.
Two federal lawsuits filed within four days of each other in California have accused Meta and OpenAI of operating outside legal frameworks that the rest of American industry treats as settled. Meta’s case, filed July 14, alleges that the company used AI-powered scoring systems to select layoff targets in a process that structurally discriminated against workers on protected leave. Apple’s case, filed July 10, alleges that OpenAI ran a coordinated campaign to steal trade secrets as the AI company built a consumer hardware division.
Algorithmic liability
Twenty-six Meta employees filed suit on July 14 in Oakland federal court, alleging that the company used AI systems to select among the roughly 8,000 workers marked for layoff in May. The complaint identifies the specific tools: keystroke and activity monitoring, AI token-usage dashboards, and algorithmically assisted performance rankings. These metrics cannot accumulate during a medical, parental, or family leave absence, and the lawsuit alleges that Meta did not adjust for protected leave when scoring employees for termination. All 26 plaintiffs had taken protected leave or received disability accommodations. Meta’s public statement asserts that workforce decisions were made by people, not AI.
The complaint invokes the Family and Medical Leave Act, the Americans with Disabilities Act, the Pregnancy Discrimination Act, and the Pregnant Workers Fairness Act. The plaintiffs also allege that Meta failed to test its AI selection systems for bias, as California and New York City laws now require. With separations scheduled to begin yesterday, the plaintiffs sought injunctive relief to preserve their employment pending arbitration. Their lawyers argue that the harms would be irreversible: employer-subsidized health coverage lost during pregnancy or active medical treatment, unvested equity forfeited, and immigration consequences triggered for employees whose visa status depends on continued employment.
The lawsuit also invokes disparate impact liability, a longstanding civil rights doctrine holding that facially neutral policies can violate discrimination law if they disproportionately burden a protected class. The Trump administration has ordered federal agencies to deprioritize disparate impact enforcement, and the EEOC has dropped cases as a result. The Meta plaintiffs are pursuing their claims independently of the agency, relying on state laws and private litigation rights that the federal pullback did not eliminate. The case tests whether AI-assisted employment decisions face the same scrutiny as any other selection process, even as the executive branch retreats from enforcement.
Show and tell
Apple filed its complaint on July 10 in the Northern District of California, accusing OpenAI’s Chief Hardware Officer Tang Yew Tan and former Apple engineer Chang Liu of orchestrating a systematic extraction of confidential information. Tan spent 24 years at Apple, most recently as vice president of product design for iPhone and Apple Watch. Liu allegedly exploited an authentication bug to access Apple’s internal network after his departure and failed to return a company-issued laptop loaded with proprietary technical documents. The complaint alleges that OpenAI coached departing Apple employees on how to avoid exit security procedures, circulating an internal Apple document that detailed how to prevent the immediate termination that Apple typically imposes when employees give notice.
The alleged extraction extended to OpenAI’s interview process. Apple job candidates were reportedly instructed to bring physical parts and prototypes for sessions that Tan organized, and Tan used internal Apple codenames when questioning candidates who still worked at the company. Jony Ive’s io Products, acquired by OpenAI for $6.5 billion in 2025, allegedly used Apple’s confidential metal-finishing techniques by misleading an Apple supplier into believing that it had Apple’s authorization. Apple first raised these concerns with OpenAI in February. OpenAI never responded.