Elon Musk’s testimony on the OpenAI dispute took a sharper, more personal turn on day two, as the billionaire told the court he now believes he was “a fool” to help fund the early launch of the organisation. In the same session, Musk offered a pointed account of what he says were Sam Altman’s motivations for building and operating OpenAI’s nonprofit structure—alleging that Altman sought the reputational and moral “halo effect” associated with a mission-driven nonprofit while, in Musk’s view, also enriching himself.
The remarks are part of a broader effort by both sides to explain how OpenAI’s governance and funding arrangements evolved from an ambitious research concept into one of the most influential AI companies in the world. At stake is not only the question of who did what and when, but also the deeper issue of whether the organisation’s structure aligned incentives with its stated mission—or whether it created opportunities for private gain under the cover of public purpose.
Musk’s comments, delivered in the courtroom setting, were striking for their bluntness. He did not frame his involvement as a strategic bet that later went wrong; instead, he portrayed it as a mistake he regrets. That framing matters because it signals that, at least from Musk’s perspective, the dispute is not merely about legal technicalities or contractual interpretations. It is about trust—about whether the people involved in shaping OpenAI’s early direction acted in good faith, and whether the nonprofit model was used as intended.
What Musk said about his own role
Musk’s testimony focused first on his early financial support. He described his decision to fund the launch of OpenAI as something he would not do again, characterising it as a “fool” move in hindsight. The implication is that Musk believed he was backing a particular kind of institution—one that would remain anchored to a nonprofit mission and operate with a level of restraint consistent with that status.
In court, however, Musk’s narrative is not simply regret. It is also an attempt to establish context: why he invested, what he expected to receive in return (not necessarily in money, but in governance and alignment), and how those expectations allegedly diverged over time. By emphasising his own misjudgment, Musk may be trying to show the court that his involvement was motivated by principle rather than opportunism. Yet he simultaneously uses that admission to sharpen the contrast between what he says he believed and what he claims actually happened.
That tension—between personal accountability and allegations of others’ motives—runs through much of the testimony. Musk’s position appears to be: even if he made a mistake by funding the effort, the organisation’s subsequent evolution should not be excused as an inevitable outcome. If the nonprofit structure was meant to serve a public mission, Musk argues, then the people running it should have been held to that standard.
Altman’s alleged “halo effect”
The most consequential portion of Musk’s day-two testimony concerned Sam Altman’s intentions and the incentives created by OpenAI’s nonprofit framework. Musk described what he called a “halo effect”—a term that, in this context, suggests the public perception and moral authority that comes with being associated with a nonprofit dedicated to safety, research, and broad societal benefit.
According to Musk, Altman sought that halo effect while, in Musk’s view, benefiting personally. The allegation is not merely that Altman gained wealth or professional status—many leaders in the tech industry do—but that the nonprofit identity functioned as a kind of shield or amplifier. Musk’s argument, as characterized in the testimony, is that the nonprofit structure could attract trust, legitimacy, and goodwill, while the individuals running the organisation could still capture value through compensation, influence, and the eventual commercialisation pathways that often accompany frontier AI development.
This is where the case becomes more than a dispute about personalities. It becomes a test of how courts interpret organisational design and incentive structures. Nonprofits are typically expected to prioritise mission over private gain. But modern technology ecosystems complicate that expectation. Frontier AI requires enormous capital, and the path from research to deployment often involves partnerships, revenue streams, and corporate entities. The question the court must grapple with is whether OpenAI’s structure was a pragmatic bridge to fund dangerous and expensive research—or whether it became a mechanism that allowed private enrichment under a nonprofit banner.
Musk’s “halo effect” framing suggests he believes the latter. He is essentially arguing that the nonprofit label carried persuasive power that could be leveraged to secure support and credibility, even if the operational reality increasingly resembled a conventional high-stakes corporate race.
Why the nonprofit structure is central
OpenAI’s organisational setup has been a focal point in the dispute because it sits at the intersection of mission and money. A nonprofit can signal commitment to public benefit, but it also raises questions about governance: who controls decisions, how conflicts of interest are managed, and how the organisation ensures that its leadership remains accountable to its stated purpose.
In Musk’s portrayal, the nonprofit structure was not just a legal formality. It was a strategic asset—one that could generate trust and legitimacy. If the court accepts Musk’s characterization, then the nonprofit framework may be viewed as something more than a vehicle for research funding. It could be interpreted as a governance arrangement that, intentionally or not, created conditions where mission language and personal incentives drifted apart.
Altman’s side, by contrast, is expected to argue that the nonprofit model was designed to ensure safety and long-term responsibility, while still enabling the scale required to compete and build. The defence narrative, as suggested by the ongoing examination in court, is likely to emphasise that the organisation’s evolution reflected the realities of AI development and that any personal compensation or career advancement was consistent with leadership responsibilities rather than a misuse of nonprofit status.
The court will therefore need to evaluate not only what was done, but why it was done—and whether the actions align with the promises implied by the nonprofit structure.
A fast-moving story with high stakes
The case is unfolding quickly, and each day’s testimony adds new layers to a dispute that already touches some of the most sensitive issues in AI governance: transparency, accountability, and the relationship between safety commitments and commercial incentives.
For observers, the courtroom drama is also a proxy for a larger societal question. As AI systems become more powerful, the institutions building them face increasing scrutiny. Governments and the public want assurances that these organisations are not simply chasing profit, prestige, or market dominance. Nonprofit structures, in theory, offer a way to embed safeguards and mission discipline. But critics argue that nonprofits can still be captured by the same incentive dynamics that shape for-profit companies—especially when the technology requires massive investment and when leadership roles carry significant leverage.
Musk’s testimony, particularly his emphasis on the “halo effect,” resonates with a broader scepticism about whether mission-driven branding can coexist with personal enrichment. Even if the legal arguments hinge on specific facts and documents, the underlying theme is familiar: when the stakes are existential, the public expects moral clarity; when the incentives are complex, the public worries that moral language can be used to justify outcomes that primarily benefit insiders.
What makes Musk’s approach distinctive
Musk’s day-two testimony stands out not only for its content but for its rhetorical strategy. Calling himself a “fool” for funding OpenAI does two things at once. First, it humanises him and suggests he is not trying to portray himself as a flawless visionary. Second, it positions his critique as coming from someone who genuinely believed in the mission enough to put money behind it—then felt betrayed by how the organisation evolved.
That combination can be persuasive to a court because it frames Musk’s allegations as rooted in disappointment rather than detachment. It implies that he is not merely attacking rivals; he is contesting a perceived breach of expectations.
At the same time, Musk’s self-criticism may also be a tactical move. By acknowledging his own error, he reduces the risk that the court sees him as purely self-serving. It also allows him to focus attention on the alleged motivations of others—particularly Altman—without appearing to claim that his own involvement was entirely rational and faultless.
The “halo effect” claim, in that sense, becomes the centre of gravity. It is the accusation that the nonprofit identity was used to create a moral aura that could obscure incentive misalignment. Whether that claim holds up will depend on evidence presented and how the court interprets intent, communications, and the evolution of governance.
The broader implications for AI governance
Even for readers not following every legal detail, the case offers a window into how AI organisations manage the tension between mission and momentum. The public often imagines that nonprofit AI labs operate like academic institutions: slow, careful, and insulated from market pressures. But the reality of frontier AI is different. Models require compute, talent, and iterative experimentation at a scale that can resemble industrial production. That scale invites partnerships and funding structures that blur the line between research and business.
When those lines blur, governance becomes the battleground. Who decides priorities? How are risks assessed? What happens when the organisation’s survival depends on commercial viability? And crucially, how do leaders ensure that personal incentives do not distort mission outcomes?
Musk’s testimony touches these questions indirectly. His allegations suggest that the nonprofit structure may have been used to maintain legitimacy while allowing the organisation to pursue paths that benefited individuals. Altman’s side, conversely, is likely to argue that the organisation’s structure was necessary to achieve its goals and that leadership compensation and influence were part of building a world-class institution.
The court’s eventual findings could influence how future AI organisations design their governance. If the nonprofit model is seen as vulnerable to incentive capture, founders may rethink structures, add stronger oversight mechanisms, or adjust how value is distributed. If the court finds that the nonprofit framework was used appropriately, it could reinforce the idea that mission-driven institutions can still scale without abandoning their principles.
What happens next
As the testimony continues, the case will likely move from broad narratives to more granular disputes: specific decisions, timelines, and the meaning of internal communications. Musk’s day-two statements set