According to Gizmodo, former OpenAI chief scientist Ilya Sutskever revealed in a recent deposition that he spent over a year planning to remove CEO Sam Altman before orchestrating his firing on November 17, 2023. Sutskever submitted a 52-page memo to the board describing Altman as exhibiting “a consistent pattern of lying, undermining his execs, and pitting his execs against one another,” specifically alleging Altman undermined CTO Mira Murati and created conflict between Sutskever and research director Jakub Pachocki. The coup succeeded briefly, but backfired when 738 employees signed a petition threatening to leave if Altman wasn’t reinstated, leading to his return on November 21, 2023. Sutskever also revealed that OpenAI and Anthropic had serious merger discussions during Altman’s brief absence, which collapsed when the board members who supported the merger stepped down. This new insight into one of tech’s most dramatic leadership crises raises fundamental questions about power and governance in the AI industry.
Sponsored content — provided for informational and promotional purposes.
The Fundamental Governance Problem
The OpenAI board structure that enabled this crisis represents a structural flaw in how we govern transformative technologies. When Sutskever and other board members attempted to remove Altman, they were operating under OpenAI’s original non-profit governance model designed to prioritize safety over profits. However, the immediate employee revolt revealed that this theoretical governance structure had completely divorced from operational reality. The court documents show how quickly idealistic governance models collapse when faced with commercial success and talent mobility. This wasn’t just a personality conflict—it was a fundamental mismatch between the board’s theoretical authority and the company’s practical power dynamics.
The Anthropic Merger That Almost Was
The revelation that OpenAI and Anthropic seriously discussed a merger during Altman’s brief absence is perhaps the most significant untold story. Had this merger proceeded, it would have consolidated two of the three leading AI labs under one roof, fundamentally reshaping the competitive landscape. Anthropic, founded by former OpenAI researchers concerned about safety, represented a completely different philosophical approach to AI development. The fact that board members were “largely supportive” of such a radical move suggests how deep the concerns about Altman’s leadership ran. This wasn’t just about replacing a CEO—it was about potentially merging OpenAI’s technical capabilities with Anthropic’s safety-first methodology.
The New Reality of Employee Power in AI
The employee revolt that forced Altman’s reinstatement reveals a fundamental shift in power dynamics within elite AI companies. When 738 employees threatened to leave, they weren’t just making empty threats—they knew their specialized skills gave them unprecedented leverage. In traditional industries, board decisions typically stand regardless of employee sentiment. But in AI, where a few hundred researchers represent the majority of the world’s expertise in large language models, employee collective action can override corporate governance. This creates a dangerous precedent where technical talent becomes the ultimate authority, potentially undermining any meaningful oversight or safety mechanisms.
Pattern of Behavior or Personality Conflict?
Sutskever’s allegations about Altman’s management style—”pitting execs against one another” and “providing different stories to different people”—echo similar concerns that reportedly led to Altman’s departure from Y Combinator. If accurate, this suggests a consistent pattern that transcends any single organization. However, we must also consider whether these management approaches, while controversial, might be precisely what enabled OpenAI’s rapid growth and product deployment. The tension between collaborative leadership and competitive drive often defines successful tech companies, and what one person calls “pitting execs against each other” another might call “fostering healthy competition.”
Broader Implications for AI Governance
This episode demonstrates why effective AI governance requires more than theoretical oversight structures. The rapid reinstatement of Altman with a new board shows how quickly idealistic governance can be replaced by practical commercial interests. As AI companies transition from research labs to commercial enterprises, we’re seeing a recurring pattern where safety-focused governance gets sidelined in favor of growth and competition. The real concern isn’t whether Altman should or shouldn’t have been fired—it’s that the mechanisms designed to ensure responsible AI development proved so fragile when tested.
What This Means for OpenAI’s Future
The aftermath of this failed coup has fundamentally reshaped OpenAI’s trajectory. The company has completed its transition to a for-profit structure and is reportedly preparing for an IPO, moves that would have been much less likely under the original governance model. The real question is whether the current leadership structure, with Altman firmly in control and the original safety-focused board dissolved, can balance the competing demands of commercial success and responsible AI development. The speed of this transformation from idealistic non-profit to commercial powerhouse suggests that when push comes to shove, market forces will consistently override governance ideals in the AI industry.
