Ex-OpenAI CTO Murati Says She Could Not Trust Altman

Mira Murati, the former Chief Technology Officer of OpenAI, testified in a videotaped deposition that she could not trust statements made by CEO Sam Altman, delivering what may be the most damaging insider testimony yet in the ongoing legal battle between Elon Musk and OpenAI. Murati stated under oath that Altman had made false representations to her regarding the safety review process for a new AI model.

The testimony emerged during the high-profile trial in which Musk alleges that OpenAI has abandoned its founding nonprofit mission in pursuit of commercial profits. Murati's words carry extraordinary weight — she was one of the most senior technical leaders at the company and briefly served as interim CEO during Altman's dramatic ouster and reinstatement in November 2023.

Key Takeaways From Murati's Testimony

• Murati explicitly stated she 'could not trust' Altman's verbal representations while he served as CEO
• The distrust centered on safety review procedures for a new AI model
• Altman allegedly made 'false representations' about how safety checks were being conducted
• The testimony was delivered under oath via videotaped deposition
• Murati departed OpenAI in September 2024 after serving as CTO for several years
• The deposition is part of Elon Musk's broader lawsuit challenging OpenAI's corporate transformation

Safety Review Deception Allegations Strike at OpenAI's Core Promise

The specific nature of Murati's allegations is particularly explosive because AI safety has been the cornerstone of OpenAI's public identity since its founding in 2015. The organization was established with the explicit mission of ensuring that artificial general intelligence benefits all of humanity, and rigorous safety protocols were supposed to be non-negotiable.

Murati's claim that Altman misrepresented the safety review process for a new model suggests a potential gap between OpenAI's public commitments to safety and its internal practices. While the specific model in question has not been publicly identified, the timeline of Murati's tenure and departure places the alleged incident during a period of rapid model development at OpenAI, including work on GPT-4o, o1, and other advanced systems.

This is not the first time concerns about safety processes at OpenAI have surfaced. In May 2024, several former employees — including co-founder Ilya Sutskever and safety researcher Jan Leike — departed the company, with Leike publicly stating that 'safety culture and processes have taken a back seat to shiny products.' Murati's sworn testimony now adds a C-suite voice to those concerns.

The Musk vs. OpenAI Lawsuit: A Battle Over AI's Future

Elon Musk's lawsuit against Sam Altman and OpenAI has become one of the most consequential legal battles in the technology industry. Musk, who co-founded OpenAI and contributed over $50 million to the organization in its early years, alleges that Altman orchestrated a betrayal of the nonprofit's original mission by steering it toward a capped-profit structure and a close partnership with Microsoft, which has invested approximately $13 billion in the company.

The lawsuit contends that:

• OpenAI was founded as a nonprofit to develop AI for the benefit of humanity
• Altman systematically converted the organization into a for-profit entity
• Microsoft's massive investment created conflicts of interest that undermined OpenAI's mission
• Key technology and intellectual property developed under the nonprofit structure is being exploited for private gain
• The planned corporate restructuring would complete OpenAI's transformation into a fully commercial enterprise

Murati's testimony strengthens Musk's narrative by providing firsthand evidence from a senior insider that the company's leadership — specifically Altman — could not be relied upon to provide truthful information even to other top executives. If a CTO cannot trust what the CEO tells her about safety processes, the argument goes, then the governance structures meant to protect the public interest are fundamentally broken.

Murati's Complex Relationship With OpenAI Leadership

Mira Murati joined OpenAI in 2018 and rose to become one of the most visible figures in the AI industry. She played a central role in the development and launch of ChatGPT, which became the fastest-growing consumer application in history after its November 2022 debut. Her technical credibility and public profile made her a critical figure in OpenAI's story.

Her relationship with the company's leadership became publicly complicated in November 2023, when OpenAI's board of directors abruptly fired Sam Altman. Murati was named interim CEO — a role she held for just 2 days before the board reversed course and reinstated Altman following intense pressure from employees, investors, and Microsoft.

Reports at the time suggested Murati had played a role in the events leading to Altman's firing, though the exact dynamics remain disputed. What is clear is that by September 2024, Murati chose to leave OpenAI entirely. In her departure announcement, she described the decision as driven by a desire to 'create the time and space to do my own exploration.' Her testimony now suggests the departure may have been motivated by deeper concerns about trust and governance.

Industry Context: Trust Deficit Growing in AI Leadership

Murati's testimony arrives at a moment when trust in AI company leadership is under intense scrutiny across the industry. The rapid commercialization of large language models has created enormous financial incentives that critics argue are overwhelming safety commitments.

Several parallel developments underscore this tension:

• Anthropic, founded by former OpenAI executives who left over safety concerns, has positioned itself as a safety-first alternative while also raising billions in venture capital
• Google DeepMind has faced internal debates about balancing research integrity with competitive pressures from OpenAI and others
• The U.S. Congress and European Union are both advancing AI regulation frameworks partly in response to concerns that companies cannot be trusted to self-regulate
• Multiple former OpenAI employees have signed open letters warning about insufficient safety practices at leading AI labs
• OpenAI's own internal safety team, the Superalignment group, was effectively disbanded in 2024 after leadership conflicts

The broader pattern suggests that the AI industry is experiencing a governance crisis. As models become more powerful and the stakes of deployment grow higher, the gap between safety rhetoric and safety practice is becoming increasingly difficult to ignore.

What This Means for OpenAI's Corporate Restructuring

The timing of Murati's testimony is particularly significant because OpenAI is in the midst of a controversial corporate restructuring. The company has announced plans to convert from its unusual capped-profit structure — where investor returns are limited to 100x their investment — to a more traditional for-profit corporation, potentially a public benefit corporation.

This restructuring has faced opposition from multiple directions. California Attorney General Rob Bonta has been investigating whether the conversion properly protects the nonprofit's assets. Musk's lawsuit directly challenges the legality and ethics of the transformation. And now Murati's testimony provides ammunition for anyone arguing that OpenAI's leadership cannot be trusted to manage this transition in good faith.

If the court finds Murati's testimony credible, it could have several practical consequences. Judges may impose additional oversight requirements on the restructuring process. Regulators may demand greater transparency about safety protocols. And investors — OpenAI has been seeking a valuation of $300 billion in its latest funding round — may face harder questions about governance risks.

Looking Ahead: The Trial's Implications for the Entire AI Industry

The Musk v. OpenAI trial is expected to continue for several more weeks, and Murati's testimony is likely just one of many revelations that will emerge. Other former and current OpenAI employees may be called to testify, and internal documents could shed further light on the company's decision-making processes.

Several key questions remain unresolved:

What specific model was involved in the alleged safety review misrepresentation? How did other senior leaders respond when they learned of the alleged discrepancy? Did the safety concerns Murati identified lead to any changes in OpenAI's review processes? And will Altman himself address the allegations directly in his own testimony?

Beyond the courtroom, the trial is setting precedents that will shape AI governance for years to come. The fundamental question — whether organizations developing the most powerful AI systems can be trusted to prioritize safety over profits — has no easy answer. But Murati's willingness to testify under oath that she could not trust her own CEO's statements about safety brings that question into sharp, uncomfortable focus.

For developers, businesses, and users who depend on OpenAI's technology, the implications are sobering. Trust is the foundation of any technology ecosystem, and when the people closest to the technology express doubts about the integrity of its stewardship, the entire ecosystem takes notice. The outcome of this trial will not only determine the future of OpenAI but may well define the standards of accountability for the AI industry as a whole.