In 2019, fast-food chain McDonald’s fired CEO Steve Easterbrook over a personal relationship with an employee. It was not a crime, but the relationship broke company policy.
In January 2023, the US Securities and Exchange Commission (SEC) charged Easterbrook with misleading shareholders about the circumstances of his exit. He settled, paying $400,000 and accepting a five-year officer and director bar, without admitting or denying the findings.
Jensen Huang now finds himself in the same boat. The NVIDIA co-founder resigned as CEO on November 12 after acknowledging an affair that reports link to Alice Huang (no relation), a former executive at Singapore-based data centre company Megaspeed International. US officials reportedly suspect her of working for Chinese intelligence. That has not been established, and China has denied it.
Subsequently, The New York Times reported that Huang also developed a relationship with Kimberly Powell, NVIDIA’s VP of Healthcare, who reported to him. Later, Reuters reported that he had an intimate relationship with a second employee.
Nothing in the available record establishes wrongdoing by NVIDIA. However, Shivaarti Bajaj, Co-founding Managing Partner at RSD Bajaj Global Law Firm, tells AIM that the legal concern “is not the existence of a personal relationship itself, but whether it influences corporate decisions, compromises fiduciary duties, or results in preferential treatment.”
There are four legal tests in play.
Did internal workplace conduct violate corporate policies, and who at NVIDIA knew what and when? Did Huang make misleading statements? Did the board exercise proper oversight under corporate law?
The fourth, antitrust, is not related to the scandal.
What Investors Must Prove
A proposed class action filed in federal court in San Jose, California, accuses NVIDIA and Jensen of misleading shareholders by failing to disclose their relationship with Alice Huang. It covers anyone who bought NVIDIA stock between October 9, 2025, and November 8, 2026, and seeks unspecified damages. The shares fell 4.3% on the day it was filed and are down about 32% from their October 6 record.
The complaint targets public statements in which an NVIDIA spokesperson and Jensen described his dealings with Megaspeed as routine. That fits the law. In Macquarie Infrastructure v. Moab Partners (2024), the US Supreme Court held that silence alone is not actionable under Rule 10b-5(b), though a half-truth is.
Mark Kistenmacher, the lead plaintiff in the NVIDIA case, is represented by The Rosen Law Firm, a plaintiff-side practice. Laurence Rosen, Founder and Partner at the firm, tells AIM that shareholders must prove four things: a false or misleading statement; that the defendants knew, or were reckless in not knowing, it was misleading; that investors relied on it; and damages. That is the test under Section 10(b) of the Securities Exchange Act of 1934.
Jensen could also be personally liable, Rosen says, for his own statements and as a control person under Section 20(a), which reaches executives who direct a company. The consequences are already visible, he says: a lower share price, Jensen’s exit, and possible congressional hearings and other government action.
NVIDIA has faced disclosure questions before. In 2022, it paid a $5.5 million SEC penalty over how it described crypto-mining demand for its gaming chips, without admitting or denying the findings.
The Board Under Scrutiny
The second suit, Vanegas v. Huang, is a derivative complaint that alleges that the company was harmed by the people meant to protect it, and shareholders sue on its behalf.
The complaint alleges breach of fiduciary duty, violations of Section 14(a) of the Exchange Act, unjust enrichment and improper insider stock sales. It says the board missed warning signs about Jensen until the matter became public, and questions whether the board was independent enough to investigate him. The independence claim has a purpose: a derivative plaintiff must either ask the board to sue or show that asking would be futile.
Oversight claims like this trace to In re Caremark (1996), a Delaware ruling, and Marchand v. Barnhill (2019), in which the Delaware Supreme Court said directors must make a good-faith effort to monitor mission-critical risks. NVIDIA is incorporated in Delaware.
A board that was misled has a defence. One that ignored warning signs does not.
Reuters earlier reported that Jensen misled several senior leaders, including General Counsel Tim Teter, about the nature of the relationship.
Bajaj says the primary safeguard when an executive’s relationship may create a conflict is timely and transparent disclosure to the board or an independent committee.
The board said at the outset that it had no reason to believe national security or NVIDIA’s secrets were compromised. It has not published findings.
On NVIDIA’s Responsibilities
Megaspeed is the centre of the export-control question. The Commerce Department opened an investigation into it in late 2024 over suspected diversion of NVIDIA chips to China. In August 2025, The New York Times reported that Teter emailed investigators asking about the probe, a move some investigators saw as potentially compromising the inquiry.
US authorities are now reportedly treating the matter as a criminal investigation, and NVIDIA has reportedly not been ruled out as a possible subject. The company has said its staff found no evidence of diversion, and Megaspeed denies moving chips illegally. Neither is a finding either way.
Export controls rest on the Export Control Reform Act of 2018 and the Export Administration Regulations, enforced by the Commerce Department’s Bureau of Industry and Security (BIS). A provider is not automatically liable because a customer diverts its products, Bajaj says, but liability can arise where it “disregards credible warning signs” or skips legally mandated due diligence. Some violations, she says, require no proof of intent.
The penalties are large. In 2023, Seagate Technology agreed to pay $300 million over hard-drive shipments to Huawei, which BIS called its largest standalone administrative penalty.
Indian operators are exposed too. AI startup Sarvam AI was allocated 4,096 NVIDIA H100 GPUs under the IndiaAI Mission, and cloud provider E2E Networks plans a 5.4 MW Blackwell expansion. An analyst earlier told Business Today that Indian cloud providers would need dedicated compliance and legal teams to clear NVIDIA’s tightening customer vetting.
The real test of accountability, Bajaj says, is not merely whether an executive had a conflict of interest, “but whether the company had effective systems to identify, disclose, and mitigate the resulting risks.”
A Second Front
The antitrust exposure predates the scandal. The US Department of Justice (DOJ) has reportedly examined NVIDIA’s acquisitions, investments, and sales practices. Reuters reports that it has also asked about the proposed $12.93 billion acquisition of open-source AI model repository Hugging Face—a deal expected to close in the first half of 2027.
The sharper question concerns Groq.
In December 2025, NVIDIA signed a non-exclusive licensing deal with AI chip startup Groq and hired its executives Jonathan Ross and Sunny Madra. Groq stays independent. The DOJ is reportedly examining whether the structure was designed to avoid merger review. Under the Hart-Scott-Rodino Act, large acquisitions must be notified to regulators before closing; a non-exclusive licence generally falls outside that requirement. When NVIDIA tried to buy Arm outright for $40 billion, it abandoned the deal in 2022 after the US Federal Trade Commission sued to block it.
The DOJ inquiry would exist had Jensen never met Alice. The securities and governance questions, perhaps, would not.






