Agentic AI
,
Artificial Intelligence & Machine Learning
,
Litigation
3-Week Court Battle Exposes Dark Side of AI Vendors and Their Promises

The jury heard closing arguments this week in the case of Musk v. Altman, and while deliberations are set to begin on Monday, no one comes out a winner in the court of public opinion.
See Also: AI Impersonation Is the New Arms Race-Is Your Workforce Ready?
Except maybe Satya Nadella.
The Microsoft CEO came across as the only grownup in the room when he testified to what he knew about OpenAI’s dealings and agreements with Elon Musk, and that he thought everything Microsoft had done to partner with the artificial intelligence company was above board. But he described the OpenAI board’s attempt to oust CEO Sam Altman in 2023 as “amateur city.”
The case couldn’t be more divisive. Musk is arguing that Altman and co-founder and president Brockman “stole a charity” when they created a for-profit arm of the nonprofit company Musk helped them get off the ground. OpenAI says Musk just has a “case of sour grapes” because he left the Silicon Valley startup just before things got good.
At the heart of this dispute is an industry with overwhelming economic power built on promises that turned out to be negotiation tactics, commitments that can be dissolved and relationships so unstable that a $13 billion investor learned that the CEO of his partner had been fired without warning.
For three weeks, the testimony coming out of the Oakland federal courthouse felt more like a teen drama than a high-stakes corporate legal dispute with nearly a trillion dollars at risk. Attorneys and their witnesses aimed to paint the opposition as liars. Dirty laundry was aired. Someone’s diary was read aloud.
The trial comes at a pivotal time for OpenAI, which is looking to launch an IPO with a valuation nearing a trillion dollars. The AI upstart soared to the top of the world economy in record time, but it came amid fierce rivalries, bad blood and maybe even some backstabbing. If Musk wins, he wants to see OpenAI rolled back into a nonprofit and for it to return as much as $150 billion to the nonprofit, essentially crippling the company.
For enterprise technology leaders one question should be top of mind as the trial comes to a conclusion: Should you be betting the house on any of these frontier AI companies?
The AI industry’s biggest players have positioned themselves as resting on a foundation of trust. Trust that you can believe what they say about their mission, values, models and accountability. Trust that the organizations now embedded in everything from your finance team to your medical records are being run by people who mean what they say.
OpenAI’s charitable mission is to create AI to benefit all humanity. Anthropic has published its lengthy “constitution” for Claude, detailing how the company hopes the model will “embody the best in humanity.”
They position themselves as the good guys – that you’ll be signing contracts with stable, well-governed organizations that operate in good faith with the best of intentions. That they are keeping the bad guys from getting their hands on the most powerful AI models the world has ever known, and that they’re doing it because it’s the right thing to do.
But several witnesses in the Musk v. Altman trial called Sam Altman a liar under oath. They were his colleagues. His former chief technology officer. His co-founder. His former board. Other reports had chronicled these rivalries and accusations, but now they’re official – on the record.
It certainly paints a challenging picture for enterprises looking to make long-term infrastructure decisions, and the tremendous commercial pressures to turn a profit are going to become even more weighty as the AI firms near IPO.
CIOs are not passive observers here. They are active players and all-important buyers in the AI ecosystem. CIOs and other enterprise leaders need to enter these contracts – whether it’s an Azure OpenAI Service contract, a ChatGPT Enterprise deployment or Claude Code – with eyes wide open. And with the same kind of thorough due diligence and contingency and redundancy planning that they put into other enterprise software and infrastructure decisions.
Earlier this year, I interviewed experts on how companies should plan to navigate the increasingly tense geopolitical environment, and they overwhelmingly agreed that one of the keys to survival was redundancy, redundancy, redundancy.
In the case of AI, that means looking into a multi-model strategy, because even in a best-case scenario, AI is moving fast. It’s pervasive, and it’s messy. If the trial has demonstrated anything to enterprise customers, it’s that these organizations are flying by the seat of their pants, playing it fast and loose with the facts, and not afraid of breaking things.
Don’t let your IT program be one of the casualties.
