← Musk v Altman Daily

OpenAI Mission Under Fire as Judge Reins In Musk (May 08, 2026)

May 08, 2026 · 6m 4s · Listen

OpenAI's mission is on trial — and so is Elon Musk's talent for staying in his lane. Welcome to Musk v Altman Daily — I'm Devin, and this courtroom has delivered the whole package: chaos, distrust, and a judge who is clearly over Elon's antics. I'm Cassidy. Today we've got testimony from OpenAI's former tech chief, a judge putting guardrails on Musk's courtroom conduct, and a witness angle nobody really saw coming — so let's get into it. If you care about who actually gets to shape the future of AI, this is the week the curtain started to come down — stay with us. Sergio Quintana, writing in NBC Bay Area:

The jury heard from three witnesses -- a former OpenAI employee working on safety issues, a board member who voted to fire Altman, and an expert on nonprofit governance -- each suggesting that the company was not properly fulfilling its mission.

Big credit to NBC Bay Area's Sergio Quintana for staying on the courthouse beat. Day eight of Musk v. Altman, and Thursday's testimony all pointed in the same direction: OpenAI has not been living up to its charitable mission. A former safety employee, a board member who voted to fire Altman, and a nonprofit governance expert — yeah, that is basically a PowerPoint deck titled 'We Told You So.' Legally, Musk has to show OpenAI materially walked away from its charitable purpose — that's the core claim. These witnesses help build the record, but sympathetic testimony is not the same thing as a winning legal theory. Rosie Campbell said she came to OpenAI because she believed in the mission, and she landed on a team literally called the Readiness Team, focused on AGI risk. If even that team got sidelined, that's a bad look. Here's Firstpost:

A former OpenAI technology chief testified on Wednesday in Elon Musk’s lawsuit that CEO Sam Altman sowed distrust among top executives as the company forged ahead with developing and broadly deploying its powerful artificial intelligence software.

Firstpost and Reuters are both on this one. Mira Murati, OpenAI's former CTO — and the woman who briefly took the CEO seat after Altman's board ouster in 2023 — testified by video that Altman was telling different people different things and, quote, 'creating chaos' at the top of the company. So the person who was literally put in as CEO to replace him is now on the stand calling him a chaos agent — that's not some annoyed middle manager, that's about as close to the inner circle as it gets. Legally, Musk needs to show OpenAI's mission was corrupted from the inside. Murati's testimony about deception and distrust helps that story, but bad management isn't automatically a legal violation. Sure — but 'he lied to my face while running the most powerful AI lab on earth' is the kind of thing that makes a jury, or a judge, start squinting at every corporate document they put in front of them. From Hadas Gold at ABC17NEWS:

Zilis’ testimony on Wednesday shed light on the fundamental role she played in the flow of information between Musk and OpenAI during critical time periods. She effectively served as a conduit between Musk and OpenAI’s leaders until Musk started his own AI company in 2023, evidence presented in court this week indicated.

Shivon Zilis took the stand Wednesday, and her testimony cuts both ways. She was an OpenAI board member, a Musk employee, and secretly the mother of his twins, all at the same time. That's a lot of hats for one courtroom. So she was basically the human equivalent of a shared Slack channel between Musk and OpenAI leadership. And nobody in that room had the whole picture. Quick update on the Zilis thread we've been following: this new testimony locks her in as the conduit between Musk and OpenAI before xAI ever went public. That's the timeline that matters most for Musk's breach-of-trust argument. She dropped off as a co-plaintiff before trial, but she did not disappear. She dropped as a plaintiff but stayed on as a witness — that's either very cooperative or very complicated, probably both. And OpenAI is going to point to her testimony and say Musk was inside the tent long enough to know exactly what was going on with the for-profit shift. From New York Post:

OAKLAND, Calif. — The judge running the bombshell trial over the future of OpenAI has apparently had it with Elon Musk’s favorite line of the last three days – “You just can’t steal from a charity” – at one point reminding Musk that he was “not a lawyer.”

Day three on the stand for Elon Musk in Oakland, and Judge Yvonne Gonzalez Rogers has officially had enough of his catchphrase — 'you just can't steal from a charity' — so much so that she struck it from the record and told him, quote, 'You're not a lawyer.' I mean, he's not wrong on the merits — but a courtroom is not a press conference, and a judge is definitely not a podcast host. That's the tension here. Legally, 'you can't steal from a charity' is a conclusion, not an argument — and Savitt's cross is built to expose that gap. Musk keeps reaching for the bumper sticker when the judge wants the receipts. Got thoughts on today's episode, a story we should be tracking, or a correction? Send us a note at muskvaltmandaily at lantern podcasts dot com. We read every message.

We've put links to every story from today's episode in the show notes, so if something caught your ear, you can follow it there and read a little deeper.

That's Musk v Altman Daily for this Friday, May 8th. This is a Lantern Podcast.