On 12 November 2026, Nvidia announced that its founder, Jensen Huang, had resigned as chief executive, president and a member of the board, with chief financial officer Colette Kress taking over as interim CEO. Huang had led Nvidia for more than three decades since co-founding it in 1993 and had become almost synonymous with the company. His sudden departure therefore prompted a strong response from employees, investors and the wider technology industry.
Of particular interest was the public support from some Nvidia employees who wanted him to stay. WIRED reported that an internal post asking the board to consider employees’ views had attracted more than 1,600 comments, most supporting his return as CEO. Claims have also circulated online that nine out of ten employees supported his remaining in charge, but such a figure cannot be treated as a formally verified vote of the entire workforce. A purported collective letter demanding the board’s resignation and Huang’s reinstatement has likewise not been reliably authenticated; its wording closely resembles the OpenAI employee letter of 2023.
Even if 90 per cent of employees, or indeed every employee, wanted Huang back, that would not alter the basic principles of governance in a listed company.
At a listed company such as Nvidia, the board has the legal authority to appoint and remove the CEO. Directors are generally elected by shareholders in accordance with the company’s governing documents and the relevant voting rights. Employees who also own shares may exercise their rights as shareholders, but being an employee does not in itself confer a vote on board membership or the appointment or removal of the CEO. Even shareholders generally do not appoint the CEO directly; they exercise oversight through the board.
Collective opposition from employees therefore does not automatically deprive a board of its legitimacy or directly overturn a decision it has lawfully made.
Employees are, of course, entitled to express their views and to decide whether they wish to continue working for the company. The board must also take seriously the damage that the departure of key employees could cause. But employees’ right to express a preference and the board’s lawful authority to appoint and remove executives operate at different levels.
There is a more significant background to this episode.
Huang has publicly acknowledged being unfaithful in his marriage, and media reports have also alleged intimate relationships with company employees. The precise nature of those relationships, whether the required internal disclosures were made, whether the company’s code of conduct was breached and whether any actual abuse of power occurred still need to be established through an independent investigation.
A distinction is necessary between private conduct and corporate governance. An extramarital relationship may seriously damage someone’s personal reputation, but it does not automatically create a legal requirement to remove them as CEO. The position changes substantially if it involves an employee over whom the executive has managerial or evaluative authority, particularly where there is deliberate concealment, a conflict of interest or unfair treatment.
A CEO controls important decisions about personnel, remuneration, resources and business strategy. An undisclosed intimate relationship with an employee affected by that authority may undermine internal fairness. Even without proven favouritism, an unmanaged conflict can give other employees reasonable grounds to question the fairness of decisions and expose shareholders to additional risk.
In a listed company, shareholders entrust management of the business to directors and senior executives, expecting them to act in the company’s interests and within its rules. If the person at the top can use personal prestige to bypass those rules, the company’s institutions may cease to impose meaningful constraints.
Nvidia’s bylaws give its board the power to remove officers through the prescribed procedures without first having to prove unlawful conduct. The existence of that power, however, is different from a legal rule requiring dismissal for a particular act. Any specific response should rest on the facts, the investigation, company rules and the relevant legal responsibilities.
This is why collective employee support cannot provide an exemption from the rules.
If the investigation ultimately establishes serious misconduct warranting disciplinary action, the board must exercise independent judgement. It cannot abandon the rules simply because Huang has achieved extraordinary commercial success or enjoys the support of most employees.
A damaged personal reputation can gradually be repaired through taking responsibility, changing one’s behaviour and demonstrating that change over time. Control of a listed company’s business should not become a means of rehabilitating an individual’s reputation. Past achievements deserve recognition, but they do not oblige a company to expose shareholders to the risks of returning someone to power.
From this perspective, a large employee campaign to bring back a founder does not necessarily protect the company’s value.
Employees may have complex motives. Years of working under a successful founder can create deep affection and identification with that person. Employees may also fear that losing a central leader will change the company’s strategy, cause further falls in its share price or affect their own jobs and incomes. Collective action that appears to express loyalty and affection may therefore also reflect anxiety about an uncertain future.
These reactions are understandable, but that does not make the underlying judgement correct.
Asking the board to explain its decision, conduct a fair investigation or properly assess the operational consequences of the founder’s departure is an ordinary exercise of employees’ right to express their views. But if serious misconduct has been established and employees still believe that a founder’s contributions exempt him from the rules, that may indicate a dangerous organisational culture.
Such a culture may itself harm shareholders. Investors already bear the risks of technological competition, changing markets and business decisions. If personal authority also comes to outrank the rules of governance, the company lacks dependable internal constraints. Investors have reason to question whether the board can supervise management independently and enforce its rules when similar problems arise.
That may increase governance risk and weaken investor confidence. Share prices, however, also reflect market conditions, business prospects, the leadership transition and other news. A fall cannot simply be attributed to an employee petition.
Some online commentators have compared Huang’s departure with Steve Jobs’s exit from Apple and eventual return. There are important differences.
Apple had already gone public in December 1980. In 1985, Jobs and the company’s then CEO, John Sculley, became embroiled in a serious conflict over the Macintosh business, corporate strategy and management authority. The board ultimately supported Sculley. Jobs lost his operational authority and subsequently left Apple to establish NeXT.
In 1997, Jobs returned through Apple’s acquisition of NeXT and again assumed a leadership role. Apple was then in serious business difficulty, and the technology, product direction and approach to management that Jobs brought back helped create the conditions for its subsequent revival.
Jobs’s departure centred on business strategy and management control. It was not a disciplinary response to an established serious conflict of interest or breach of professional conduct. The success of his eventual return therefore does not establish that Huang should be offered the same opportunity.
Another useful comparison is the OpenAI board’s sudden dismissal of Sam Altman as CEO in November 2023, followed by his reinstatement.
OpenAI’s governance structure at the time differed markedly from Nvidia’s. Although it had commercial operations, it was controlled by the board of a non-profit organisation. That board’s mission was different from the shareholder-centred governance arrangements of an ordinary listed company.
Altman’s role also had particular significance. OpenAI was at a critical stage in the rapid development of large language models. Training those models required not only algorithms and researchers, but also substantial funding, GPU computing capacity, data centres and infrastructure relationships with partners such as Microsoft. Altman played an important part in organising research and development, raising capital, coordinating commercial partnerships and securing computing resources.
OpenAI already had about 770 employees at the time; it was not a small company comprising just a handful of key people. Nevertheless, its research capabilities, access to funding and crucial partnerships remained highly dependent on its leadership team.
After Altman’s dismissal, more than 700 employees signed a letter saying they might leave OpenAI and join the new team Microsoft had offered to establish for him unless the board changed. Microsoft also indicated that it was willing to take those employees. The dispute consequently became a crisis over whether OpenAI could retain its existing organisation and research capacity.
That went beyond employees simply expressing dissatisfaction. They had no direct authority to appoint a CEO, but they could choose to leave. If the overwhelming majority of core researchers departed together, the original organisation could lose its capacity to continue its research. Its name and legal entity might survive while its technical team and organisational capabilities moved to another platform.
There was also another important factor in the OpenAI episode. An independent review completed in 2024 found that Altman’s dismissal had principally arisen from a breakdown of trust between him and the board. Although the board had the authority to dismiss him, it had acted on an abbreviated timetable, without a full inquiry or an appropriate opportunity for him to respond. The review also found that his conduct did not mandate removal.
Altman’s return therefore cannot be reduced to a victory for collective employee pressure. Problems with the board’s own process and judgement, together with the practical risk of losing the core team, also shaped the outcome.
The experiences of Jobs, Altman and Huang may look similar on the surface, but the reasons for their departures, their companies’ governance structures, their organisations’ stages of development and the legal risks differ substantially. Where serious breaches of professional conduct have been established, the fact that another company once rehired its founder cannot simply be treated as a precedent.
The three cases do, however, reveal a deeper shared issue: a company’s dependence on its founder or another central leader.
Over more than three decades, Huang helped transform Nvidia from a graphics processor company into a major force in AI computing. His technical judgement, management ability and industry relationships made an undeniable contribution to its success.
But a mature listed company cannot base its continued existence on the assumption that one individual will always be able to do the job.
Even without a controversy over personal conduct, illness, an accident, retirement or some other unforeseen event could suddenly prevent Huang from continuing to lead the company. The board should already have considered those possibilities, developed a credible succession process and ensured continuity in research, product planning, customer relationships and management.
The appointment of Colette Kress as interim CEO demonstrates some capacity to manage an emergency handover. A temporary appointment, however, is different from a mature long-term succession arrangement. Whether Nvidia has a leadership team capable of assuming the founder’s responsibilities over the longer term will have to be tested in practice.
If a listed company established more than thirty years ago still cannot find successors able to take on its founder’s core responsibilities, that exposes a structural risk deserving serious attention, however remarkable that founder’s past achievements may be.
Such risks can build over time. As a founder repeatedly succeeds, employees and executives become accustomed to relying on that person’s judgement. A board impressed by a long record of strong performance may also neglect the development of successors or fail to build management arrangements appropriate to the company’s growing scale.
The result is a business that expands in technology, markets and financial strength while its internal governance continues to revolve heavily around one person.
An early exposure of that dependence need not be entirely negative.
If the board uses this episode to address conflicts of interest, clarify the limits of the chief executive’s authority, develop a new leadership team and establish a more reliable succession process, short-term disruption could lead to better long-term governance.
Nvidia’s commercial and technical capabilities do not automatically disappear with Huang’s departure. The company has accumulated substantial technology, talent, intellectual property, product systems and customer relationships. What now needs to be tested is whether those resources can continue to work effectively under a new leadership structure.
A founder may be exceptionally capable and even difficult to replace at a particular stage in a company’s development. But if, after more than thirty years, the business remains highly dependent on one person, the board has grounds to examine its own performance.
For a listed company, the value of institutions lies precisely in their ability to apply the same basic principles of oversight, accountability and succession even to its most authoritative and accomplished leader.
A founder can create a great company. Whether that company can continue to succeed after the founder leaves is an important test of the maturity of its organisation and governance.
—
References
Nvidia’s announcement of its CEO transition, 12 November 2026; Nvidia’s executive transition filing with the US Securities and Exchange Commission; WIRED’s report on employees seeking Huang’s return; Reuters’ analysis of potential legal liability; Historical information on Apple’s public listing; The collective OpenAI employee letter of 2023; OpenAI’s published findings from the 2024 independent review.
Facts and sources checked as at 14 November 2026. The specific facts and responsibilities concerning alleged breaches of Nvidia’s internal rules remain subject to the findings of the formal investigation.
Discover more from Geoffrey Chen
Subscribe to get the latest posts sent to your email.