Elon Musk Concludes Testimony in OpenAI Lawsuit Alleging Misuse of Charitable Assets

On April 30, 2026, Elon Musk stepped down from the witness stand in a federal courtroom in Oakland, California, after three days of testimony that ranged from wistful recollection to open hostility. His target: OpenAI, the artificial intelligence company he co-founded in 2015 as a nonprofit research lab, and which has since grown into an $852 billion juggernaut [1]. The central allegation is straightforward, even if the legal terrain is not — Musk claims that OpenAI's leadership converted charitable assets into private wealth, betraying the organization's founding mission [2].

The trial, presided over by U.S. District Judge Yvonne Gonzalez Rogers, is not merely a dispute between two tech billionaires. Its outcome could reshape the legal framework governing how nonprofits commercialize their work, set binding precedent for AI governance, and determine the fate of more than $130 billion in assets [3].

The Founding and the Money

OpenAI was incorporated in December 2015 as a 501(c)(3) nonprofit, with Musk and Sam Altman as co-chairs. The stated mission: develop artificial general intelligence (AGI) that would be "safe and broadly beneficial" to humanity, as a counterweight to the concentration of AI capabilities at Google [4].

Musk's financial contributions to the organization have been a contested point. During testimony, Musk characterized his total giving as approximately $44 million, a figure largely consistent with OpenAI's own disclosure that his contributions were "below $45 million" [5]. Court records show at least $15 million in directly traceable donations, supplemented by quarterly payments of roughly $5 million for about one year, plus office space in San Francisco [5].

A pivotal moment came during cross-examination by OpenAI attorney William Savitt on day three of the trial. Musk acknowledged that no written agreement or contract existed specifying the terms under which his donations were made [6]. There was no document stipulating that the money had to remain within a nonprofit structure, nor any formal governance rights tied to his contributions. "At the end of the day, you can't steal a charity," Musk said in response, framing his argument as one about fiduciary duty rather than contractual rights [7].

The Conversion: From Nonprofit to $852 Billion Corporation

In 2019, one year after Musk resigned from the board citing a potential conflict with Tesla's own AI work, OpenAI created a "capped-profit" subsidiary — OpenAI LP. The structure capped investor returns at 100x their investment, with any surplus flowing back to the 501(c)(3) [8]. Microsoft invested $1 billion that same year, the beginning of a partnership that would eventually total more than $13 billion [3].

Source: Visual Capitalist / Sacra

Data as of Apr 30, 2026CSV

The valuation trajectory tells a story of exponential growth. From $1 billion in 2019 to $852 billion by April 2026, OpenAI's market value exploded following the release of ChatGPT in late 2022 [1]. As the company grew, so did pressure to restructure further. In October 2025, OpenAI completed a recapitalization that converted the for-profit subsidiary into a public benefit corporation (PBC), with the newly created OpenAI Foundation retaining a 26% stake — valued at approximately $130 billion — and Microsoft holding 27% [9].

Musk's complaint identifies this conversion as the moment charitable assets crossed a legal line. His legal team argues that under California law, when a nonprofit converts or dissolves, its assets must be distributed "to another charity with the same or similar purposes" [10]. The complaint alleges that instead of honoring this principle, OpenAI's leadership engineered a structure that enriched insiders — particularly Altman and President Greg Brockman — while dressing up the transaction in mission-preserving language [2].

What California Law Actually Says

California's Nonprofit Public Benefit Corporation Law (Corporations Code §§ 5110–6910) imposes strict requirements on the disposition of charitable assets. When a nonprofit converts, dissolves, or merges with a for-profit entity, those assets cannot simply become private equity. The Attorney General has supervisory authority to ensure compliance [10].

California Attorney General Rob Bonta investigated OpenAI's restructuring over roughly 18 months. Through negotiations, Bonta's office secured several concessions: that charitable assets would be directed to their intended purposes, that safety would be prioritized, and that OpenAI would remain headquartered in California [11]. With those conditions in place, Bonta declined to oppose the recapitalization.

But critics argue the deal was insufficient. Public Citizen, a consumer advocacy group, calculated that the nonprofit's controlling interest alone was worth at minimum $30 billion, based on a standard 20% control premium applied to OpenAI's then-$150 billion valuation [12]. The group sent letters to both the California and Delaware attorneys general demanding that any conversion payments go to an independent charitable enterprise, not back to an OpenAI-affiliated foundation [12].

The fact that the Attorney General signed off does not preclude the court from reaching a different conclusion. Musk's lawsuit seeks remedies including reversion of OpenAI to its nonprofit structure, removal of Altman and Brockman from the board, and up to $130 billion in damages directed to OpenAI's charitable foundation [3].

The Competitor Question

OpenAI's defense rests heavily on one premise: Musk is not a disinterested philanthropist. He is the owner of xAI, a direct competitor that launched in 2023 [13].

During cross-examination, Savitt pressed Musk on this point repeatedly. Musk acknowledged that xAI competes with OpenAI but characterized it as "much smaller" with "very small market share" [13]. The defense argues that the lawsuit is a competitive weapon — an attempt to hobble a rival through litigation rather than innovation.

The most damaging moment for Musk's credibility on this front came on day three, when he admitted under oath that xAI had "partly" used distillation — a technique in which one AI model is trained by querying another's outputs — to build its Grok chatbot using OpenAI's models [14]. Distillation is not explicitly illegal, but it violates OpenAI's terms of service [14]. The admission undercut Musk's positioning as a principled defender of OpenAI's mission, supplying the defense with ammunition to argue the lawsuit is about market advantage.

Still, the legal theory underlying Musk's case does not depend on his motives. Breach of charitable trust and unjust enrichment claims are evaluated on the merits of the alleged misconduct, not the plaintiff's reasons for bringing the action [15]. Several legal scholars have noted that even if Musk's competitive interests are transparent, the underlying question — whether OpenAI's conversion violated charitable trust law — remains a legitimate and unresolved area of law [15].

Source: Polymarket / Kalshi

Data as of Apr 30, 2026CSV

Prediction markets as of late April 2026 reflect genuine uncertainty. Polymarket priced Musk's odds of winning at 43%, while Kalshi traders gave him a 53.5% chance [16]. Only 8% of traders expected a settlement exceeding $10 billion [16]. The divergence across platforms suggests that even informed observers find the legal outcome genuinely difficult to predict.

What the Internal Documents Show

Discovery in the case has surfaced internal communications that both sides have cited as supporting their narrative. OpenAI published a timeline of emails and messages, including ones in which Musk himself acknowledged the need for a for-profit structure to raise the capital necessary for AI research [9]. The defense argues these communications show Musk was aware of and initially supportive of the structural shift before souring on it after losing influence.

Musk's legal team, however, points to other internal discussions showing that OpenAI's leadership was aware as early as 2018 that the conversion could create tension with the original nonprofit charter [15]. The precise content of these documents has been the subject of intense courtroom argument, with Judge Gonzalez Rogers warning lawyers on both sides that "AI itself isn't on trial" and urging them to stay focused on the specific legal claims [17].

The Stakeholder Fallout

The trial's outcome has material consequences for a broad ecosystem. OpenAI employs roughly 3,000 people, many of whom hold equity in the for-profit entity [3]. Microsoft, with $13 billion invested and a 27% stake in the PBC, is a named defendant — Musk's complaint alleges Microsoft aided and abetted the alleged breach of charitable trust [3]. SoftBank invested $40 billion, half of which was reportedly conditional on OpenAI lifting its profit cap and allowing traditional equity by early 2026 [8].

If Musk's requested remedies were granted in full — reversion to nonprofit status and $130 billion in damages — the consequences would be severe. Employee equity could be wiped out. Microsoft's investment structure would require fundamental renegotiation. SoftBank's conditional investment terms would be thrown into question. None of these entities have publicly disclosed contingency plans for such an outcome [3].

How OpenAI's Conversion Compares

OpenAI's nonprofit-to-commercial transition has few direct precedents, though several analogies have been drawn.

Mozilla Foundation created the Mozilla Corporation in 2005 to manage the Firefox browser commercially while the Foundation retained its public-interest mission. Unlike OpenAI, Mozilla did not convert charitable assets into private equity — the Corporation was established as a new entity, with the Foundation retaining control [18].

Patagonia moved in the opposite direction. In 2022, founder Yvon Chouinard transferred all voting stock to the Patagonia Purpose Trust and all nonvoting stock to the Holdfast Collective, a 501(c)(4) focused on climate work [19]. This was a for-profit becoming mission-controlled, not a nonprofit becoming commercial.

Neither case involved the scale of capital at issue in the OpenAI dispute, nor did either trigger comparable regulatory scrutiny. The OpenAI conversion is, in practical terms, an experiment without a clean historical parallel — which is precisely why the trial's outcome will function as new law.

Implications for Other AI Organizations

If Musk's legal theory prevails, the ripple effects would extend well beyond OpenAI. Anthropic, the AI safety company founded by former OpenAI researchers, operates as a for-profit under supervision of a Delaware purpose trust that can appoint a majority of its board [20]. While Anthropic was never a nonprofit, its governance structure is designed to prioritize mission over profit — an arrangement that could face new legal scrutiny if courts establish a strong precedent around the enforceability of mission commitments.

DeepMind, acquired by Google in 2014, originally operated with an ethics board that was supposed to govern the application of its research. The board's actual authority has been debated, and DeepMind has been fully integrated into Google's corporate structure [20].

If the court rejects Musk's theory — finding that OpenAI's conversion was lawful and adequately compensated the nonprofit — the precedent could be equally significant. It would signal that mission-driven organizations can transition to commercial structures provided they follow the procedural requirements and secure regulatory sign-off, even if the original donors object. That outcome could encourage other AI safety nonprofits and research labs to pursue similar hybrid paths with reduced fear of legal challenge.

What Comes Next

Musk's testimony has concluded, but the trial is expected to continue for several weeks. Sam Altman is widely expected to testify — Polymarket prices the probability at 85.5% [16]. Microsoft executives may also be called.

The nine-person jury will serve in an advisory capacity; the final decision on remedies rests with Judge Gonzalez Rogers [3]. The legal questions — whether charitable trust obligations follow assets through corporate restructuring, whether the Attorney General's sign-off immunizes the conversion from judicial review, and whether a co-founder's donations create enforceable trust obligations absent a written agreement — are all matters of first impression at this scale.

The case is also unfolding against a charged political backdrop. OpenAI itself filed a request in early April asking the California and Delaware attorneys general to investigate what it called Musk's "anti-competitive behavior" [21]. Judge Gonzalez Rogers dropped Musk's fraud claims at his request on April 24, just before trial, while allowing the core breach-of-charitable-trust and unjust enrichment claims to proceed [16].

Whatever the verdict, the Musk v. OpenAI trial has already accomplished something rare: forcing a public accounting of how the most valuable AI company in the world went from a nonprofit research lab to an $852 billion corporation — and whether that transformation was legal.