Elon Musk helped write the mission. Sam Altman's company is now being sued for abandoning it. And in a separate courtroom, families are alleging OpenAI valued its IPO over their children's lives.
This is the paradox at the heart of two legal assaults converging on OpenAI this week. In one, Musk—the company's most prominent co-founder—is testifying in federal court that OpenAI's pivot to for-profit structures betrayed its founding promise to develop artificial intelligence safely and for humanity's benefit. In the other, seven families of victims from the Tumbler Ridge school shooting in Canada are suing OpenAI and CEO Sam Altman for negligence, alleging the company suppressed warnings about the shooter's violent conversations to protect its public offering.
The two cases share no plaintiffs and no legal theory. But they orbit the same accusation: that OpenAI chose commercial interests over its stated mission.
In San Francisco, Musk's testimony has centered on emails and documents showing his heavy influence over OpenAI's original nonprofit structure. Evidence revealed this week includes exchanges with Jensen Huang, whose early gift of an in-demand Nvidia supercomputer helped launch the organization. Musk has argued this infrastructure—and his credibility as a co-founder—was weaponized to transform the company into something its charter never intended.
OpenAI's defense has focused on the evolving nature of AI development, arguing that the original nonprofit model became untenable as the compute requirements for competitive research exploded. The company contends it remains committed to safety even as it pursues commercial viability.
The Tumbler Ridge case presents a starker moral calculus. The families allege that OpenAI's systems identified Jesse Van Rootselaar's conversations about gun violence, that company executives discussed alerting police, and that they ultimately decided against it. The Wall Street Journal reported OpenAI considered the flag but chose silence—reportedly to shield its IPO prospects.
If true, this isn't a governance dispute. It's an accusation that OpenAI weighed a child's life against a stock ticker and chose the latter.
OpenAI has not confirmed these specific allegations and the case remains in early stages. But the company's own acknowledgment that it considered action—before declining to take it—suggests the system worked, at least partially. The question becomes whether consideration without action constitutes negligence.
Both cases could reshape AI industry liability. A ruling against OpenAI in Musk's governance claim would force restructuring of countless hybrid nonprofit-for-profit arrangements. A ruling for the Tumbler Ridge families would establish that AI companies have affirmative duties to act on safety signals—and that failure to do so carries legal consequences.
For the families, the stakes are concrete and irrefutable. For Musk, they're philosophical and financial. For OpenAI, they may determine whether the company survives as currently structured. The question both juries will ultimately answer is simple: what did OpenAI owe the world, and did it pay that debt?
A verdict is expected in both cases before the end of the year.