Can AI Leaders Be Trusted?
The Musk-OpenAI trial wrapped up this week with closing arguments, but according to TechCrunch AI’s Equity podcast, the jury isn’t the only one weighing a verdict. The bigger question hanging over the proceedings has nothing to do with whether OpenAI broke its nonprofit promises. It’s whether anyone running a frontier AI lab can be trusted at all.
TechCrunch AI reports that Musk’s attorney Steve Molo spent the final days grilling Sam Altman about congressional testimony where Altman claimed he had no equity in OpenAI. Technically, his stake came through Y Combinator as a passive VC investment. Altman tried to wave it off by saying “everybody understands what it means to be a passive investor in a VC fund.” Molo’s response landed: do you really think the congressman questioning you knew that?
That moment is doing a lot of work. It’s not about a single semantic dodge. It’s about the entire posture of an industry that asks the public, regulators, and investors to take its word on existential questions.
Two styles of not-quite-truth
What stands out in the TechCrunch AI coverage is the contrast between the two principals. Musk has a documented history of tweeting things that turn out to be false, then walking them back combatively on the stand. Altman went the opposite direction: affable, conflict-averse, “I’m working on it.”
Kirsten Korosec put it bluntly on the podcast: both men have been untruthful, but their handling of it diverges sharply. Sean O’Kane went further. “I don’t trust him,” he said of Altman. “But you know, I don’t trust most people, so I guess that’s just the baseline.”
This is the part worth sitting with. Tech reporters who cover these companies daily are openly questioning whether the CEOs running them tell the truth. That’s not a courtroom problem. That’s an industry problem.
Why this lands now
Korosec zoomed out on the podcast and named the real issue: “This is a fundamental question for a lot of tech journalists, policymakers, and more and more consumers, about all the AI labs. It’s really come down to trust, because we don’t have the insight. These are all privately held companies, there’s a lot behind the veil still.”
That veil matters. Frontier labs are asking for:
- Hundreds of billions in capital commitments
- Regulatory frameworks built around their own safety claims
- Public acceptance of products trained on contested data
- Government contracts and infrastructure access
All of it depends on taking the labs at their word about capabilities, safety practices, and intentions. There’s no audited disclosure. No public filings. No SEC oversight until an IPO that may never come.
What practitioners should take from this
A few practical implications:
- For enterprise buyers: Trust-but-verify isn’t a slogan anymore. Get contractual commitments on data handling, model behavior, and uptime. Verbal assurances from any lab CEO won’t hold up.
- For investors: The governance discount on private AI companies is real and growing. Structures that worked for SaaS don’t map onto labs sitting on dual-use technology.
- For policymakers: Voluntary commitments have a credibility problem. Mandatory disclosure regimes will accelerate, especially in the EU and likely in select US states.
- For founders building on top of these labs: Diversify your model providers. The Musk-Altman fight is a preview of what happens when platform risk meets personality risk.
What comes next
The jury verdict matters less than the cultural shift the trial revealed. As O’Kane noted, “all these people came out of this looking a little bit worse.” That’s the takeaway. Musk wanted to sling mud at a rival. He may have succeeded, but the mud spattered everywhere, including on the broader narrative that AI leaders deserve the benefit of the doubt.
IPOs would force disclosure, but neither OpenAI nor Anthropic seems in a hurry. Until then, the gap between what these companies say and what outside parties can verify keeps widening. The trial didn’t cause that gap. It just made it visible.
For the full TechCrunch AI conversation, including the Equity podcast breakdown, see the original coverage.