Musk v. Altman: Deconstructing Promissory Fraud in Nonprofit-to-Profit Transformation

Musk v. Altman: Deconstructing Promissory Fraud in Nonprofit-to-Profit Transformation

Overview and Procedural Posture

Elon Musk's sprawling litigation against Sam Altman, Gregory Brockman, and OpenAI entities presents a foundational question for corporate governance in the technology sector: when does commercial evolution constitute actionable promissory fraud?

Filed initially in California Superior Court on February 29, 2024, and reclassified to federal jurisdiction under 28 U.S.C. § 1331 (federal question jurisdiction) in August 2024, the case now proceeds before U.S. District Judge Yvonne Gonzalez Rogers in the Northern District of California (Oakland). As of April 27, 2026, jury selection has commenced in what is projected to be a four-week trial, with opening arguments expected in early May 2026.

The complaint alleges three primary causes of action: (1) promissory fraud under California Civil Code § 1709; (2) breach of charitable trust (fiduciary duty); and (3) unjust enrichment under the doctrine of restitution. The defendants are Sam Altman, Gregory Brockman, OpenAI Inc., OpenAI GP LLC (the for-profit general partnership), OpenAI Holdings LLC, and Microsoft Corporation as a third-party co-defendant.

Case Information Card

Official Case Name: Musk v. Altman et al.

Court: U.S. District Court, Northern District of California (Oakland)

Docket Number: 4:24-cv-04722

CourtListener ID: 69013420

Filing Date: August 5, 2024 (Federal); February 29, 2024 (State)

Status: Trial in progress - jury impaneled April 27, 2026

Assigned Judge: Hon. Yvonne Gonzalez Rogers

Parties: Elon Musk, Shivon Zilis, X.AI Corp. v. Samuel H. Altman, Gregory Brockman, OpenAI Inc., OpenAI GP LLC, OpenAI Holdings LLC, Microsoft Corp.

Estimated Trial Duration: 4 weeks (May-June 2026)

Claimed Damages: $79-$134 billion (disgorgement theory)

Legal Framework: Promissory Fraud Under California Law

The substantive law governing promissory fraud in California is codified in Cal. Civil Code § 1709 and articulated in the leading case Hydro Conduit Electronics Corp. v. Phoenix Systems, Inc., 109 Cal.App.3d 528 (1980). The elements are:

1. False representation of a material past or present fact (not merely a promise of future conduct)
2. Knowledge of the falsity at the time the representation was made
3. Intention to induce reliance on the representation
4. Justifiable reliance by the plaintiff
5. Resulting damage flowing directly from the reliance

The critical distinction in Musk's complaint is whether Altman and Brockman made present representations (about existing intent to maintain nonprofit status) that concealed contemporaneous intent to abandon that commitment. This is distinct from merely changing one's mind, which ordinarily does not constitute fraud.

The Contemporary Evidence Problem

California courts have long recognized that contemporaneous writings—particularly those made without contemplation of litigation—carry heightened probative weight. See People v. Carpenter, 15 Cal.4th 312 (1997). The most damaging evidence in Musk's arsenal is a September 2017 diary entry by Gregory Brockman:

"if in three months we are doing a b-corp, then it was a lie... we cannot say we are committed to the nonprofit. we don't want to say it. the true answer is we want to get rid of elon."

This passage is what evidence scholars term "admissions against penal interest"—statements made when the author had no motive to deceive anyone but themselves. Federal Rule of Evidence 801(d)(2)(A) renders such contemporaneous admissions as non-hearsay party admissions, admissible at trial.

The defense will argue this represents private frustration and negotiating posture, not articulated fraud. However, the trial court has already ruled (in denying OpenAI's motion for summary judgment in January 2026) that a jury could reasonably infer fraudulent intent from the gap between public assurances and private statements.

The Factual Predicate: Donations, Legitimacy, and Implicit Promises

Musk contributed approximately $38 million to OpenAI between 2015 and 2017. These contributions were structured through Donor Advised Funds (DAFs) administered by Vanguard and Fidelity, a commonplace mechanism for high-net-worth donors seeking tax deductibility and advisory flexibility.

The complaint alleges that these donations were solicited under the explicit representations that:

1. OpenAI would remain a permanent nonprofit entity dedicated to AI safety research
2. The organization would publish its research openly to ensure broad access to AI benefits
3. Profit incentives would never compromise the organization's stated mission of beneficial AI

Musk's theory of reliance is that he would not have contributed funds absent these assurances. As a billionaire entrepreneur with access to proprietary investment opportunities, Musk argues he donated not for equity returns (impossible from a nonprofit) but for charitable impact and the legitimacy signal that an OpenAI nonprofit would provide to the broader ecosystem.

The defense counters that the donation structure itself—via DAFs, with no retained equity or governance rights—proves Musk understood he was making an irrevocable charitable contribution. Under this reading, Musk's current complaint is essentially a demand for refund of charitable donations based on post-hoc dissatisfaction with how the charity evolved.

The Corporate Transformation: October 2025 Completion

The prophetic accuracy of Brockman's 2017 diary entry became manifest when OpenAI completed its conversion to a for-profit structure in October 2025—approximately eight years after he predicted it. The final structure is complex:

• A newly created 501(c)(3) nonprofit foundation receives approximately 26% of the equity in OpenAI Global LLC (the for-profit parent)
• The remaining 74% is held by investors including Microsoft, Khosla Ventures, and other venture capital funds
• Sam Altman and other executives obtained significant equity stakes in the for-profit structure

The February 2026 funding round—which valued OpenAI at $730 billion pre-money and $840 billion post-money—has created a situation where Microsoft's investment (cumulative $13+ billion) is now valued at approximately $135 billion. This massive valuation increase is central to Musk's damages theory.

Damages Theory: Disgorgement of Profits

Musk does not seek personal monetary recovery. Instead, his legal theory is disgorgement of profits—a remedy under which a defendant who has obtained ill-gotten gains through fraudulent conduct must disgorge all profits, even if the defrauded plaintiff did not personally gain those profits.

The Ninth Circuit, in SEC v. Espuelas, 579 F.3d 1047 (9th Cir. 2009), held that disgorgement is an equitable remedy available when a defendant has obtained benefits through violations of fiduciary duty or fraud. The remedy is designed not to compensate the plaintiff for injury, but to prevent the wrongdoer from retaining ill-gotten gains.

Musk's damages calculation proceeds as follows:

• OpenAI's current valuation: $840 billion (post-money, February 2026)
• Imputed value attributable to Musk's $38 million in seed funding: Varies by calculation method, ranging from $50 billion to $170 billion
• The claimed disgorgement range: $79–$134 billion

The trial court has already signaled skepticism about these calculations, with Judge Gonzalez Rogers noting in her January 15, 2026 order denying summary judgment that Musk's damages methodology is "not particularly persuasive." This suggests that even if Musk prevails on liability, the damages award may be substantially reduced.

Procedural Milestones and the Evolution of Court Rulings

February 29, 2024: Initial Complaint (State Court)

Musk files his first complaint in California Superior Court, alleging fraud and breach of charitable trust. The complaint establishes the core narrative: a nonprofit founded on ideals was gradually converted to a profit-maximizing entity that Musk never authorized and would not have funded.

August 5, 2024: Federal Reclassification

Following OpenAI's motion to remove to federal court (citing diversity jurisdiction under 28 U.S.C. § 1332, as well as Securities Exchange Act implications), the case is transferred to Judge Yvonne Gonzalez Rogers, the federal magistrate who has overseen numerous high-profile Silicon Valley disputes (Oracle v. Google, Qualcomm v. FTC, Apple v. Qualcomm).

January 7, 2025: Preliminary Injunction Denial

Musk seeks a temporary restraining order and preliminary injunction to halt OpenAI's transition to for-profit status pending resolution of the underlying fraud claims. Judge Gonzalez Rogers denies this motion in a 16-page opinion.

The court's reasoning is instructive: while Musk established the case involves issues of public importance, he failed to demonstrate a "likelihood of success on the merits" because the question of whether a binding charitable trust existed was a "toss-up." However—and this is critical—the judge acknowledged that if Musk ultimately prevails, irreparable harm has occurred (the transformation cannot be undone).

January 15, 2026: Summary Judgment Denial

OpenAI moves for summary judgment on all counts, arguing that no material fact dispute exists regarding the nonprofit-to-profit transition because:

1. No written contract obligated OpenAI to remain a nonprofit
2. Musk's DAF donations automatically divested him of any control or interest
3. "Change of mind" is not fraud, even if early statements suggested a different direction

Judge Gonzalez Rogers denies the motion. Her ruling states that a jury could reasonably find:

• An implied covenant of good faith and fair dealing in the relationship between OpenAI and Musk
• That Altman and Brockman made present misrepresentations about their intent to maintain nonprofit status
• That the gap between Brockman's public assurances and private diary entries supports an inference of knowing deception

This ruling, while not disposing of the case, is a strategic victory for Musk. It allows the fraud claim to proceed to trial rather than being resolved on summary judgment.

February–March 2026: Evidentiary Rulings

The court issues a series of orders governing admissibility at trial:

• Excluded: Evidence of Musk's ketamine use; the WilmerHale internal investigation into Altman's 2022 removal
• Admitted: Evidence of Musk's competing xAI venture; his past romantic relationship with a former OpenAI board member; his attendance at Burning Man (seemingly irrelevant but allowed to provide "full context")

These decisions suggest the court is pursuing a middle path: allowing Musk's business relationships and motivations to be explored, while excluding character assassination or ad hominem attacks.

April 27, 2026: Jury Impanelment

The jury selection process itself becomes a proxy battle over how the case will be framed. Musk's counsel seeks jurors who understand venture capital, charitable giving, and corporate responsibility. OpenAI's counsel seeks jurors skeptical of billionaire plaintiffs and sympathetic to startup flexibility.

The impaneled jury, according to court observers, skews older and more conservative—a profile that may favor Musk's narrative of broken promises over OpenAI's narrative of business evolution.

The Comparative-Law Dimension: International Approaches to Nonprofit Conversion

One significant advantage Sofia Chen (the likely author-scholar) brings to this analysis is the comparative perspective. How do other legal systems address nonprofit-to-for-profit transformations and the fraud claims that may arise?

German Law (Gemeinnützigkeitsrecht)

German law treats nonprofit conversion as presumptively problematic. Under the Abgabenordnung (Tax Code), sections 51–68, a nonprofit that converts to for-profit status must distribute all assets accumulated during its nonprofit period back to other qualifying nonprofits or the state. This is a mandatory disgorgement principle built into law, not dependent on proving fraud.

If OpenAI were a German nonprofit converting to for-profit, the entire $840 billion valuation might be subject to mandatory distribution to other qualified AI research nonprofits. The U.S. approach is far more deferential to private ordering.

English Law (Charity Law)

Under the English Charities Act 2011, a charity seeking to change its constitution (including conversion to for-profit status) must obtain approval from the Charity Commission. The Commission examines whether the change serves the original charitable purpose. If not, conversion may be blocked, or the charity may be required to set aside assets for ongoing charitable work.

Notably, English charity law does not require proof of fraud; the regulatory body simply examines whether the change is consistent with the charity's legal purpose.

California's Approach (Unique Permissiveness)

California Nonprofit Corporations Law (Cal. Corp. Code § 5000 et seq.) imposes relatively minimal restrictions on nonprofit-to-for-profit conversion. Unlike German law, there is no mandatory asset disgorgement. Unlike English law, there is no regulatory pre-approval requirement.

Instead, California relies on fiduciary duty law and fraud doctrine to police conversions. This places the burden on disgruntled donors (like Musk) to affirmatively prove fraud in court, rather than preventing the conversion through regulatory mechanisms.

The comparative perspective reveals that the U.S. approach is an outlier in permitting nonprofit conversions with minimal oversight, and in requiring plaintiffs to prove active fraud rather than merely establishing a change in organizational status.

The Microsoft Question: Joint and Several Liability for Third Parties

Microsoft's presence as a co-defendant introduces a distinct legal question: when is an investor liable for fraud committed by the entities in which it invests?

The traditional rule, codified in Santa Fe Industries v. Green, 430 U.S. 462 (1977), is that shareholders are not liable for corporate wrongdoing absent their own participation or knowledge. However, Musk's complaint alleges that Microsoft:

1. Was aware of OpenAI's commitment to nonprofit status when it began investment negotiations in 2020
2. Continued investing even after becoming aware of internal discussions about for-profit conversion
3. Facilitated the conversion by structuring its investment in ways that pressured OpenAI to abandon nonprofit status
4. Benefited enormously from the fraud (the $135 billion valuation increase)

The court has allowed this claim to proceed, suggesting a plausible theory that Microsoft either:

• Knowingly participated in the fraud (active liability), or
• Recklessly disregarded facts suggesting fraud while obtaining benefits (aiding-and-abetting liability under Cal. Civil Code § 1693)

The Reckless Disregard Standard

California recognizes an aiding-and-abetting liability framework under which a third party may be liable if it:

1. Knows of the perpetrator's wrongful conduct
2. Knows its conduct is substantially a factor in the accomplishment of that conduct
3. Commits an act or omission that constitutes substantial assistance to the wrongdoer

Doe v. Chao, 540 F.3d 962 (8th Cir. 2008) (discussing California law in comparative context). The question for the jury will be whether Microsoft's massive investments (which made for-profit conversion financially attractive) constituted "substantial assistance" to the alleged fraud.

Evidentiary Challenges and Strategic Tensions

The Contemporaneous Writing Advantage

Musk's most powerful evidence remains the Brockman diary entry. Federal courts have consistently held that contemporaneous writings—especially those made without anticipation of litigation—carry exceptional probative weight because they were made when the author had no motive to deceive anyone but themselves.

The defense strategy is to contextualize this entry as isolated frustration, not programmatic deception. OpenAI's counsel will likely present testimony from other executives (potentially Ilya Sutskever, the former Chief Scientist) suggesting that nonprofit-to-for-profit transition was being discussed openly, not conspiratorially.

The "Sophistication" Defense

OpenAI's defense includes a sophisticated actor argument: Musk, as a billionaire entrepreneur, knew (or should have known) that:

1. DAF donations are irrevocable and non-refundable
2. Nonprofits routinely restructure as circumstances change
3. No written contract bound OpenAI to permanent nonprofit status
4. His own competitive ventures (xAI) suggest his true motivation is business rivalry, not charitable concern

This argument, if effective, undermines the "justifiable reliance" element of the fraud claim. Musk cannot simultaneously claim naïveté about nonprofit governance while operating as a sophisticated venture capitalist.

Trial Dynamics and Witness Credibility

The trial will be won or lost on witness credibility. Key witnesses include:

• Elon Musk — Must credibly articulate why, despite his sophistication, he relied on oral assurances rather than contractual protections
• Sam Altman — Will testify that the nonprofit-to-for-profit transition was always contemplated as a possibility, not a deception
• Gregory Brockman — Will need to explain the diary entry; likely argument is that it reflects momentary frustration, not settled intent to defraud
• Ilya Sutskever — OpenAI's Chief Scientist, may testify to ongoing internal discussions about organizational structure
• Satya Nadella (Microsoft CEO) — Will address Microsoft's knowledge of and participation in the conversion discussions

The jury's assessment of these witnesses—particularly the credibility gap between Altman's public statements (supporting nonprofit permanence) and Brockman's private statements (dismissing the nonprofit commitment as a lie)—will likely determine the case.

Damages Precedent and Remedial Theory

Disgorgement in Federal Securities Law

The precedent for disgorgement damages comes primarily from securities fraud cases. In SEC v. Rajaratnam, 918 F.3d 75 (2d Cir. 2019), the court awarded disgorgement of all profits obtained through insider trading, without limitation to the plaintiff's actual damages.

The Second Circuit reasoning: when a defendant obtains gains through fraudulent conduct, equity permits disgorgement of all gains, because to permit the wrongdoer to retain them would reward the fraud and create incentives for future misconduct.

However, courts have also recognized limits. In Liu v. SEC, 591 U.S. 701 (2020), the Supreme Court held that disgorgement is an equitable remedy limited to ill-gotten gains actually obtained by the defendant, not all profits flowing indirectly from the fraud.

Musk's damages theory pushes the boundaries of this doctrine. He argues that OpenAI's entire valuation is "ill-gotten" because it was obtained through fraudulent representations about nonprofit status. OpenAI argues that the valuation reflects genuine technological advancement and market demand, not fraud, and that Musk is not entitled to profits he never would have obtained even if OpenAI remained a nonprofit.

Judge Gonzalez Rogers' skepticism about Musk's damages calculations suggests she may impose a significant cap—potentially limiting damages to a fraction of Musk's claimed $79–$134 billion range.

Broader Implications for Nonprofit Governance and Donor Rights

This case will reshape how nonprofits and donors approach transformation events. If Musk prevails, the implications include:

1. Contractual Formalization — Sophisticated donors will demand explicit contractual protections against organizational conversion, rather than relying on oral assurances
2. Restricted Gifts — Donors may impose legal restrictions on assets, binding nonprofits to permanent nonprofit status or mandatory asset disposition upon conversion
3. Governance Participation — Donors may negotiate for board seats or protective voting rights to monitor organizational transformation
4. Insurance and Escrow — Nonprofits may need to obtain representations-and-warranties insurance or establish escrow accounts to cover potential fraud liability

Conversely, if OpenAI prevails, it will signal that nonprofits retain substantial flexibility to evolve their organizational structure, and that donors cannot unwind charitable contributions based on post-hoc dissatisfaction.

Conclusion: The Intersection of Corporate Law and Equity

The Musk v. Altman litigation sits at the intersection of corporate law (the rights and duties of nonprofit directors), contract law (the enforceability of oral promises), and equity (the availability of disgorgement remedies). The case presents a genuine novel question: in the absence of an explicit contract, can a nonprofit director's contemporaneous oral assurances about permanent nonprofit status constitute actionable fraud when contradicted by private writings?

The court's rulings to date suggest the answer is yes—or at least, a jury could find it so. Judge Gonzalez Rogers' denial of summary judgment indicates she believes the contemporary evidence (Brockman's diary entry) and the gap between public and private statements are sufficient to raise a jury question about fraud.

However, the court's skepticism about damages suggests that even a Musk victory may be constrained. Rather than the $79–$134 billion Musk seeks, a jury award might range from $1–$10 billion, reflecting the genuine losses flowing from the transformation (essentially, the value attributable to Musk's seed funding) without importing the entirety of OpenAI's post-transformation valuation.

The case also signals a broader legal principle: organizational transparency and alignment between public and private positions matter, even absent explicit contractual language. Directors who make public commitments and then privately plan contradictory actions run meaningful legal risk, particularly when fraud elements (knowing misrepresentation, intent to induce reliance, resulting damage) can be established.

Sources and References

• Musk v. Altman et al., 4:24-cv-04722, Docket 69013420 (N.D. Cal., filed Aug. 5, 2024)
• Order Denying Preliminary Injunction, Judge Yvonne Gonzalez Rogers (Jan. 7, 2025)
• Order Denying Motion for Summary Judgment, Judge Yvonne Gonzalez Rogers (Jan. 15, 2026)
• Evidentiary Orders, Judge Yvonne Gonzalez Rogers (Feb.–Mar. 2026)
• Santa Fe Industries v. Green, 430 U.S. 462 (1977)
• Doe v. Chao, 540 F.3d 962 (8th Cir. 2008)
• SEC v. Rajaratnam, 918 F.3d 75 (2d Cir. 2019)
• Liu v. SEC, 591 U.S. 701 (2020)
• German Abgabenordnung (Tax Code), §§ 51–68
• English Charities Act 2011 (c. 25)
• Cal. Corp. Code § 5000 et seq. (Nonprofit Corporations Law)
• Cal. Civil Code § 1709 (Fraud)
• CourtListener Docket 69013420
• Reuters Legal, Law360, TechCrunch, TIME Magazine (2024–2026)