Key Takeaways
- Court testimony alleges Sam Altman provided inconsistent information to boards and colleagues, creating an environment of chaos.
- Altman reportedly misrepresented the completion of AI safety reviews for a ChatGPT variant to the OpenAI board.
- Sam Altman’s acknowledgment of an indirect equity stake in OpenAI contradicts his earlier Senate testimony. Former colleagues and board members of OpenAI have now placed their concerns about Sam Altman’s honesty into federal court records. Testimony in the Elon Musk lawsuit against OpenAI, heard in May 2026, goes well beyond the company’s nonprofit-to-for-profit conversion: it has put Altman’s candor with boards, his handling of safety reviews and his financial disclosures under direct legal scrutiny. The picture that emerges from the proceedings is not flattering.
1. Allegations of Inconsistent Candor with Boards
The November 2023 board crisis at OpenAI is the clearest public record of this tension. The board removed Altman as CEO, citing a lack of consistent transparency, before reinstating him days later. That episode now has a formal companion in court testimony. Mira Murati, OpenAI’s former Chief Technology Officer, is reported to have testified that Altman often said “one thing to one person and completely the opposite to another person,” producing an environment of “chaos” and, at times, deception. Musk’s lawyers pressed Altman directly on these points during cross-examination, referencing prior statements from former associates.
The concerns are not confined to OpenAI. According to reports, Y Combinator founder Paul Graham told colleagues that Altman “had been lying to us all the time” before his departure from the accelerator, where he had served as president. That account has resurfaced as part of the current trial, adding to a pattern of concern that spans more than one institution.
2. Misrepresentation of AI Safety Processes
The most operationally significant allegation in the proceedings concerns AI safety reviews. According to court testimony, Altman told the board that three safety reviews for a ChatGPT variant had been completed when only one had been performed. The gap between those two numbers matters: in AI development, safety sign-offs are not administrative formalities. They determine whether a model r
Ilya Sutskever, OpenAI’s co-founder and former Chief Scientist, is reported to have sent internal memos to the board expressing concern about Altman’s honesty, specifically alleging that Altman downplayed the need for safety approvals in conversations with executives including Murati. When Murati reportedly raised these claims with OpenAI’s general counsel, the counsel’s response — that he was “confused where Sam got that impression” — and the broader interaction have been cited in legal proceedings, reflecting a breakdown in internal accountability around a process that carries genuine public-safety weight.
3. Ambiguous Financial Disclosures
Altman’s financial relationship with OpenAI became a focal point in court after he acknowledged holding a stake in a fund managed by Y Combinator, which in turn owns shares in OpenAI. During earlier Senate testimony, he had stated he held no equity directly in the company. Musk’s lawyers argued the court admission showed he had misled Congress.
The technical distinction, indirect ownership through a fund versus direct equity, is real. Whether it is a meaningful one depends on what Altman understood Congress was asking. That ambiguity is at the centre of the legal argument. For a company operating with a hybrid nonprofit and for-profit structure, and a stated mission to benefit humanity broadly, the threshold for disclosure clarity is arguably higher than for a standard private firm. The court proceedings have made that expectation explicit in a way that Senate hearings did not.
4. A Leadership Style Described as Chaotic and Deceptive
Murati’s testimony went beyond specific incidents. She described Altman as actively “creating chaos” and characterised his approach as conveying different information to different people, making it difficult for the board to reach informed decisions.
Carroll Wainwright, a former OpenAI researcher, is quoted as suggesting that Altman “sets up structures that, on paper, constrain him in the future, but then, when the future comes, and it comes time to be constrained, he does away with whatever the structure was.” That description, if accurate, points to something more systemic than individual miscommunication: a leadership pattern in which formal constraints function as performance rather than as genuine checks. The result, according to multiple former colleagues, was a working environment shaped by unpredictability at the top.
5. A Pattern of Distrust Across Multiple Ventures
The concerns raised in court are not new. Before OpenAI, Altman led the location-sharing startup Loopt, where, according to reports, senior employees became sufficiently concerned about his leadership that they urged the board to remove him, citing a lack of transparency. The Graham statement about Y Combinator followed. The late Aaron Swartz, who passed through Y Combinator in Altman’s cohort, allegedly described him in terms that will not be repeated here as fact, given they cannot be verified from a primary source.
The New Yorker’s April 2026 report drew these threads together, and the Musk trial has now placed them in a legal context. Whether the pattern constitutes disqualifying dishonesty or reflects the aggressive communication style common among Silicon Valley founders is a question the proceedings will not resolve cleanly. What the trial has done is move these accounts from the realm of industry gossip into sworn testimony and court filings, where they carry a different evidentiary weight. The governance implications for OpenAI, particularly as it navigates its structural conversion and increasing regulatory attention on AI development, including scrutiny of how AI hiring and deployment decisions are made, extend well beyond the outcome of this lawsuit.
For more coverage of AI policy and regulation, visit our AI Policy & Regulation section.
Originally published at https://autonainews.com/sam-altmans-candor-safety-and-finances-face-court-scrutiny/
Top comments (0)