Industry

Altman Testifies Musk Suggested Passing OpenAI to Kids

Sam Altman testified in the Musk v. Altman trial that Elon Musk once proposed handing control of OpenAI to his children if he died. Musk's lawyers questioned Altman on deception claims and investments, while Altman described Musk as fixated on dominating the company. The session featured tense exchanges and details on Altman's stakes in firms partnering with OpenAI.

Neura News

Neura News

Neura Market Editorial

May 13, 20265 min read

Originally reported by wired.com

Altman Testifies Musk Suggested Passing OpenAI to Kids

Altman Testifies Musk Suggested Passing OpenAI to Kids

Sam Altman appeared on the witness stand Tuesday in the Musk v. Altman trial. He faced hours of questions from Elon Musk's attorneys about claims of deceptive conduct. Altman depicted Musk as deeply focused on gaining command over OpenAI.

The questioning provided Musk's side with a key opportunity. Musk had faced challenges in building a strong argument so far. Exchanges grew intense as Altman addressed claims from past colleagues labeling him unreliable.

Presenting such points matters for swaying the jury. It also aims to shift views in public discussions. Just before the trial began, Musk sent a text to OpenAI president Greg Brockman. In it, he warned that he and Altman would soon become the most hated men in America.

Musk's suit charges Altman with taking over the OpenAI nonprofit. It claims Musk's $38 million gift to the charity funded a for-profit entity now valued above $850 billion.

Tuesday offered scant proof to fill holes in Musk's claims. Altman and Sam Teller, once Musk's chief of staff, stated they remembered no unique terms tied to Musk's contributions to OpenAI. Evidence mounts that Musk launched the suit too late. Years passed after his final donation and growing doubts about OpenAI's adherence to its mission. The statute of limitations had run out by then.

Attendees and Musk's Departure

Greg Brockman attended with his wife Anna. OpenAI chief futurist Joshua Achiam joined them in the gallery. Altman and Brockman watched Musk testify earlier. Musk left before Altman's turn. Records show he headed to the Washington, DC area Tuesday, set to travel to China with President Donald Trump.

Altman first shared his perspective under mild queries from OpenAI counsel. He wore a purple tie. He described himself as an entrepreneur and investor drawn to artificial intelligence's strength and risks.

Altman said Musk fixated on directing OpenAI for years. He cited a particularly hair-raising moment. Musk proposed control of OpenAI go to his children should he pass away. "We didn't feel comfortable with that," Altman said. He also called Musk's 2018 push for a Tesla AI division, where he offered Altman a leadership role, a vague, lightweight threat. Musk seemed poised to overpower OpenAI regardless.

Intense Cross-Examination

Steven Molo, Musk's counsel, opened sharply. He asked Altman, "Are you completely trustworthy?" Altman said he believed himself to be. Molo followed by questioning if the jury should believe Altman's prior statements. Altman replied, "That's up to them. I'm not going to tell the jury what to think."

The sharp back-and-forth continued:

Molo: Do you always tell the truth?

Altman: I'm sure there is some time in my life where I have not.

Molo: Do you tell lies to advance your business interests?

Altman: No.

Molo: Have you misled people with whom you do business?

The #1 Newsletter in AI

Stay ahead of the AI curve

The most important updates, news, and content — delivered weekly.

No spam. Unsubscribe anytime.

Altman: I do not think so.

Molo: Do they think so?

Altman: I can't answer that for other people.

Molo: You've repeatedly been called a liar by people you've done business with, right?

Altman: I've heard people say that.

Molo reviewed charges from ex-OpenAI leaders and directors. These included Musk, former chief scientist Ilya Sutskever, ex-chief technology officer Mira Murati, board member Tasha McCauley, and Anthropic founders Dario and Daniela Amodei. He pressed Altman on each. Molo even raised issues from 15 years back at Loopt, Altman's location-sharing startup. There, claims surfaced he overstated daily active users.

Altman sidestepped at times. He requested specific testimony citations and said he knew nothing of some claims.

Molo drew firmer responses on Altman's stakes in OpenAI partners. News emerged Monday night of a House oversight panel letter to Altman. It sought details on financial ties possibly clashing with OpenAI duties.

Investments and OpenAI Stewardship

Altman holds nearly a $2 billion stake in nuclear firm Helion. OpenAI agreed with Helion in 2024 for potential AI energy buys if fusion succeeds. Helion sells no power yet. Altman owns $600 million in Stripe. He invests in Reddit and Cerebras, both with OpenAI pacts.

Altman countered he guided OpenAI well. The nonprofit ranks among history's best-funded charities. It controls over $200 billion in OpenAI equity. "I do not believe I could have taken any other actions to get $200 billion into a nonprofit," said Altman.

Questions on Power and Governance

Altman portrayed Musk as power-driven. Yet Molo argued Altman shared that trait. Altman confirmed considering a California governor bid around 2017. Later that year, he publicly dropped the idea.

Molo challenged OpenAI's nonprofit board power to hire and dismiss leaders. OpenAI insists it holds that authority. Molo showed texts post-Altman's firing from Altman, Microsoft CEO Satya Nadella, and executive Bret Taylor. They picked agreeable new board picks, including Taylor as chair.

Altman said he sought those guarantees to return. The firing board chose replacements and brought him back.

"I almost didn't come back because I was so upset. I thought there had been such a failure in governance, and the way the board had gone about this left such a lack of confidence from the people that needed to run the company," said Altman. "Everything I had built was going to get destroyed. I was willing to run back into a burning building."

Altman now serves on OpenAI's board, common for CEOs, he noted. Molo asked if Altman would dismiss himself. Altman, possibly returning Wednesday, said no such plans exist.

OpenAI started in 2015 as a nonprofit to advance safe AI. Founders included Musk and Altman. Musk departed in 2018 amid control disputes. OpenAI shifted to a capped-profit model in 2019 for funding.

Related on Neura Market

More from Neura News

AI Models

Google Unveils Gemini 3.6 Flash, 3.5 Flash-Lite, and Cyber Model

Google has released three new Gemini models: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber. The 3.6 Flash model offers improved coding and knowledge work with 17% fewer output tokens and lower costs. The 3.5 Flash-Lite is the fastest in the series at 350 tokens per second, designed for high-throughput agentic tasks. The 3.5 Flash Cyber model, available only to governments and trusted partners via CodeMender, focuses on finding and fixing cybersecurity vulnerabilities. Google also noted that Gemini 3.5 Pro is being tested with partners and that pre-training for Gemini 4 has begun.

Jul 21·5 min read
AI Models

Alibaba Qwen-Image-3.0 renders infographics and tiny text in one pass

Alibaba's Qwen team released Qwen-Image-3.0, an image generator designed for practical applications like newspaper layouts and complex infographics. The model processes prompts of up to 4,500 tokens and can render legible text as small as ten pixels, mathematical formulas, and twelve languages in a single pass. It is currently available through invite-only API access, with plans to integrate it into first-party apps like Qwen Chat soon.

Jul 21·4 min read
AI Models

Google Unveils Gemini 3.6 Flash, 3.5 Flash-Lite, and Cyber Model

Google DeepMind has introduced three new Gemini models: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber. The 3.6 Flash model offers improved coding and multimodal performance with 17% fewer output tokens and lower cost. The 3.5 Flash-Lite is the fastest in its series at 350 output tokens per second, designed for high-throughput agentic tasks. The 3.5 Flash Cyber, fine-tuned for cybersecurity, will be available exclusively to governments and trusted partners via the CodeMender agent.

Jul 21·6 min read