OpenAI Trial: Musk Demands $150B — Satya Nadella Testifies

The OpenAI trial commanded Silicon Valley's attention this week as Elon Musk demands $150 billion from the AI company he co-founded — the industry's most public reckoning with its own founding promises. On Monday, May 11, Microsoft CEO Satya Nadella takes the witness stand. The week before, a woman who shares four children with Musk and once served on OpenAI's board testified as his personal AI adviser. Whatever your position on Musk, this AI trial is rewriting how the industry understands its own origins.

A $150 Billion OpenAI Lawsuit Built on a Broken Promise

OpenAI was founded in 2015 with an explicit nonprofit mission: develop artificial general intelligence (AGI — AI that can match or exceed human intelligence across all tasks) for the benefit of all humanity, not for profit. Musk was a co-founder and early backer. He left the board in 2018 amid reported strategic conflicts.

By 2023, OpenAI had deployed ChatGPT to over 100 million users and locked in a $10 billion investment from Microsoft. Musk argues this pivot to commercial products betrayed the founding mission. In his 2024 lawsuit, Musk makes four specific demands:

• Remove Sam Altman and Greg Brockman (OpenAI's president) from their roles immediately
• Stop OpenAI from operating as a public benefit corporation (PBC — a legal structure that blends nonprofit ideals with for-profit operations)
• Block a proposed restructuring that would award Altman significant equity in the company
• Award Musk up to $150 billion in damages

OpenAI's defense is characteristically blunt. The company called the lawsuit "a baseless and jealous bid to derail a competitor" — noting that Musk has been building his own AI company, xAI, and its Grok chatbot, throughout the same period. The implication is clear: this is business warfare dressed up as principled protest.

The Witness Nobody Expected: Shivon Zilis Takes the Stand

The trial's most revealing pre-Nadella testimony came from Shivon Zilis — a venture capitalist, former OpenAI board member, and — as publicly confirmed during her testimony — the mother of four of Musk's children.

She testified under oath that she served as a "Musk adviser" working on his "entire AI portfolio: Tesla, Neuralink, and OpenAI" starting in 2017. That means Zilis was simultaneously advising Musk while serving on the OpenAI board — a textbook conflict of interest (when a person holds competing loyalties that can distort their decision-making) that raises pointed questions about what information may have crossed between the two organizations during those years.

Her testimony matters because it directly speaks to Musk's claim that he remained a genuine, active stakeholder in OpenAI's mission long after his public departure from the board. If the jury credits her account, it bolsters Musk's narrative that he was a wronged insider — not simply a wealthy competitor looking to handicap a rival.

Satya Nadella Testifies May 11 — Then the Scientist Who Built GPT-4

Microsoft CEO Satya Nadella is scheduled to testify on Monday, May 11. As the architect of Microsoft's $10 billion investment in OpenAI — the largest single corporate stake — Nadella's testimony could answer some of the trial's most consequential questions:

• What did Microsoft know about the November 2023 Altman firing, and when did they know it?
• Did Microsoft play any role in Altman's reinstatement just five days after being fired?
• How did the investment agreement address governance rights, if at all?

Following Nadella: Ilya Sutskever — OpenAI's former chief scientist and one of the most consequential AI researchers in history — is lined up to testify. Sutskever voted to fire Altman in November 2023, reportedly wept about it shortly after, then signed a letter demanding Altman's return. His first-hand account of what happened inside that boardroom is potentially the trial's defining moment.

For context: when the board fired Altman in November 2023, it cited that he was "not consistently candid in his communications with the board." Within five days, Altman was reinstated, most of the original board had resigned, and the incident became a landmark case study in what happens when nonprofit governance collides with a $90+ billion commercial valuation.

OpenAI Launches a Crisis Safety Tool While Battling in Court

While its legal team battles Musk's lawsuit, OpenAI also unveiled Trusted Contact this week — a mental health safety feature allowing ChatGPT users to designate emergency contacts (a parent, friend, or therapist) who will be notified if the AI detects conversations involving self-harm or suicidal ideation.

OpenAI's stated rationale: "when someone may be in crisis, connecting with someone they know and trust can make a meaningful difference." The feature is currently opt-in (requires users to proactively configure it) for adults only — meaning it must be set up before any crisis occurs. Opt-in crisis tools historically see very low adoption rates, which limits their real-world reach regardless of intent.

The timing is not accidental. As Musk's lawsuit accuses OpenAI of abandoning its humanitarian mission, Trusted Contact lets OpenAI demonstrate — in court and in public — that it still prioritizes people's wellbeing over profit. Whether it becomes a meaningful mental health intervention depends entirely on how many users actually enable it.

Musk's Other Big Bet: A $55 Billion AI Chip Factory in Texas

Outside the courtroom, Musk is simultaneously making a historic infrastructure play through SpaceX. The company is planning "Terafab" — a semiconductor (microchip) manufacturing plant in Austin, Texas — with an initial investment of $55 billion. Additional expansion phases could push total investment to $119 billion.

For scale: the entire U.S. CHIPS Act — the 2022 federal law designed to revitalize American semiconductor manufacturing — allocated $52 billion across the entire industry. SpaceX's single Terafab facility could exceed that figure. The investment reflects Musk's intent to build vertically integrated AI infrastructure: not just AI models, but the custom silicon chips that train and run them — entirely independent of the existing NVIDIA and Microsoft supply chain.

If Terafab succeeds at scale, it would give xAI and Tesla's Dojo supercomputer a domestic chip pipeline insulated from dependence on NVIDIA and TSMC (Taiwan Semiconductor Manufacturing Company — the firm that fabricates most of the world's most advanced chips). That is both a geopolitical hedge against potential disruption and a direct assault on NVIDIA's near-monopoly on AI training hardware. If you follow AI infrastructure, Terafab is the story to watch — stay current with the latest AI news as this story develops.

Three Other AI Moves That Defined This Week

Google Quietly Killed Project Mariner on May 4

Google shut down Project Mariner on May 4, 2026 with minimal fanfare. Mariner was a browser-based AI agent (an AI that acts autonomously on your behalf — clicking, scrolling, and filling forms across the web) capable of handling up to 10 simultaneous web tasks. Google first revealed it to the public in December 2024.

The shutdown note was brief: "Thank you for using Project Mariner. It was shut down on May 4th, 2026 and its technology voyaged to other Google products." The web-task capabilities are now absorbed into Gemini Agent. This is Google's consistent pattern: test a feature as a standalone experiment, then fold what survives into Gemini. If you were using Mariner for AI automation workflows, switch to Gemini Agent now — or browse our AI automation guides to find the right tool for your use case.

Apple AirPods Are Getting Cameras — Two Production Steps Away

Bloomberg's Mark Gurman reports that Apple AirPods equipped with built-in cameras have entered "design validation test" — the stage immediately before mass production validation. These are not cameras for photos or video recording. They capture low-resolution visual data to help Siri answer context-aware questions: scan the items on your kitchen counter to suggest a recipe, identify objects in your environment, or recognize what is in front of you.

Apple's hardware-AI strategy centers on embedding sensors into devices people already carry daily, rather than pushing AI entirely to the cloud. No commercial launch date has been announced — "design validation" is still two production phases from store shelves.

Google Fitbit Air: $99, Screenless, and Deliberately Simple

Google announced the Fitbit Air at $99 — a screenless fitness tracker with a metallic fabric clasp, positioned as the modern successor to the original clip-on Fitbit One from 2012. It is not a direct Whoop competitor (Whoop targets serious athletes with a subscription-based performance monitoring model). The Air remains in design validation phase with no confirmed ship date. Watch whether Google adds AI-powered health insights before launch — or deliberately ships it simple as a counterpoint to an increasingly complex wearables market. The minimalist bet could either differentiate or misfire.

Related Content — Get Started with AI Automation | AI Automation Guides | More AI News