The difference is not a universal rule about workplace relationships. It is who controls a company, what its board knows, and which risks it decides it can no longer carry.
A famous technology founder faces scrutiny over personal relationships. His company depends heavily on his reputation. Investors wonder whether the rules that apply to ordinary employees also apply to the person at the top.
That description can sound familiar. But it does not tell you who has the power to end the founder’s tenure.
On November 12, 2026, Nvidia announced that Jensen Huang had resigned as president, chief executive and director, effective immediately. Chief financial officer Colette Kress became interim CEO while retaining her finance role. Harvey Jones became the board’s first chair.
The announcement did not explain the departure’s cause or publish the findings of the company’s investigation. That matters. “Huang resigned amid scrutiny” is supported. “Nvidia proved an affair and automatically fired him for it” goes beyond what the announcement says.
The comparison with Elon Musk raises a real question. But the answer is not that corporate America has a single misconduct rule and randomly chooses when to enforce it.
What Nvidia has actually said
The sequence began before the resignation.
In a November 8 message, Huang acknowledged being unfaithful to his wife, apologized and supported an independent investigation into reported national-security concerns. That was not an admission that he had disclosed company secrets or committed espionage.
The board’s statement that day said it had no reason to believe Nvidia’s secrets or national security had been compromised. It also expressed confidence in Huang. Those were the board’s statements at that moment, not a completed investigation’s conclusions.
On November 10, independent directors said they had retained Joon H. Kim and Wachtell, Lipton, Rosen & Katz to investigate. Subsequent reporting described an alleged relationship with an employee who reported directly to Huang. Two days after the independent directors’ announcement, Nvidia confirmed his departure.
This is a more complicated situation than an executive admitting marital infidelity. Workplace reporting relationships, possible conflicts of interest and separate security concerns can present different questions. Reporting about each should not be blended into one supposedly established offense.
A board does not need a criminal verdict to change CEOs
A corporate leadership decision is not a criminal trial.
Directors may need to weigh trust, continuity, disclosure, employee confidence and risk before a court has considered anything. Conversely, an allegation does not establish that misconduct occurred merely because a CEO leaves.
Nvidia’s 2026 proxy statement says its executive officers serve at will and do not have individual employment, severance or change-in-control agreements. Its code of conduct addresses conflicts of interest and applies to leaders as well as other employees.
Neither document establishes a universal rule that every personal relationship produces the same outcome. A relationship involving a direct report can raise questions about favoritism, disclosure and the integrity of decisions, even where the people describe it as consensual. Whether those issues occurred in a particular case requires evidence.
The public documents explain the framework. They do not reveal every consideration behind the Nvidia board’s decision.
Being the founder is not the same as controlling the vote
This is where the Musk comparison becomes useful.
A founder can be extraordinarily important without having enough voting power to control director elections. Another founder can have a smaller share of the economic ownership than the voting structure might lead you to expect, yet still hold decisive influence over the board.
SpaceX’s June 2026 IPO prospectus describes two share classes with different voting rights. It projected approximately 82.4% of post-offering voting power for Musk, with the higher-vote class entitled to elect a majority of directors. That is the prospectus’s dated figure, not a calculation of his exact ownership today.
The structural point is more important than the decimal: a company can sell shares to the public while retaining concentrated founder control.
That does not make directors irrelevant or a founder immune from law. It changes the practical setting in which challenges to leadership occur. The person whose conduct is being scrutinized may also have enormous influence over who sits on the board scrutinizing it.
Musk’s companies are not one employer
It is tempting to treat “Musk” as a single corporate institution. Tesla, SpaceX and Neuralink are separate organizations with different governance arrangements and employment relationships.
SpaceX’s voting structure cannot simply be pasted onto Tesla as an explanation of its decisions. Nor can every report about a colleague, partner or co-parent be treated as proof of the same workplace conduct.
The Wall Street Journal’s 2024 investigation reported relationships involving SpaceX employees, including a direct report. Accounts differed, and some were disputed. Those distinctions belong in the comparison. Having children through IVF with a colleague does not, on its own, prove an affair or harassment.
There is also a straightforward correction to the idea that Musk never faces consequences: his 2018 settlement with the Securities and Exchange Commission required him to leave Tesla’s chairmanship, although he remained CEO. That case concerned securities allegations, not employee relationships. It shows why “kept his job” can hide significant differences between roles and consequences.
The comparison needs a fair test
Imagine two companies with equally famous chief executives. At the first, independent shareholders can elect directors without the CEO determining the outcome. At the second, the CEO holds enough voting power to dominate that election. Both boards have responsibilities, but they operate under different practical constraints.
Now add a second difference: one board is considering a disclosed relationship, while another is investigating whether important information was withheld. Add a third: one controversy concerns only the workplace; another includes separate concerns about access to sensitive information.
Those are hypothetical distinctions, not findings about the people in this story. They show why “both had relationships with employees” would still be an incomplete comparison even if every relationship were established. The employer has to evaluate the circumstances, not simply check whether two headlines contain the same noun.
There is a similar problem with measuring consequences. Resigning as CEO, leaving the board, losing the chairmanship and surrendering voting control are different events. An executive can lose one position and retain another source of influence. A useful comparison specifies the position lost rather than describing everyone as either untouchable or completely removed.
A governance explanation is not an excuse
Explaining concentrated control does not mean endorsing every outcome it produces. Nor does a founder’s departure prove that the board followed an exemplary process.
The public can reasonably ask whether the directors were sufficiently independent, whether the inquiry was thorough and whether employees were protected. But those are questions to investigate, not conclusions supplied automatically by a resignation announcement.
Nvidia’s independent directors publicly identified the outside lawyers conducting their inquiry. That is a concrete process disclosure. It does not establish what the lawyers will conclude, which witnesses they have interviewed or whether every relevant document has been reviewed. Readers should resist treating the existence of an investigation as either a conviction or an exoneration.
Likewise, the ability to control director elections does not erase company policies, fiduciary responsibilities or legal exposure. Voting power explains a channel of influence. It is not a defense to every allegation and cannot establish what any director actually thought.
This is also why stock-market success cannot supply the missing answer. A strong business performance does not determine whether conduct occurred or whether it was acceptable. A falling share price does not prove wrongdoing. Both can affect the environment in which directors make decisions without resolving the underlying facts.
For readers trying to make sense of the next founder scandal, the useful habit is to separate three columns: verified conduct, formal consequences and the governance structure. The comparison becomes clearer when those columns are not allowed to substitute for one another.
The real question is who can hold a founder accountable
Huang’s departure and Musk’s continued leadership invite a comparison. They do not establish that two boards evaluated identical evidence under identical rules and reached opposite verdicts.
The better questions are concrete. Who appoints the directors? Who controls the votes? What was disclosed? Which company policies apply? What does an independent investigation establish? What risks are the directors responsible for managing?
Those questions are less simple than “Why him, but not Musk?” They are also where the answer lives.
For Nvidia, an important part remains unresolved: the investigation’s findings. Until those are public, ownership and governance can explain how accountability works without pretending to know everything that caused this particular resignation.
