When David Reber quietly stepped down from his role as Nvidia’s longtime chief security officer and head of product security this past August, outsiders had little reason to pay close attention. High-level executive transitions happen regularly in the fast-paced technology sector, and corporate security leadership changes rarely make mainstream financial headlines. Yet, just two and a half months later, the departure of Nvidia CEO Jensen Huang under a cloud of controversy has cast that summer transition in an entirely new light. Huang’s exit followed mounting questions regarding his personal relationships—beginning with an executive at Megaspeed, a Singapore-based data center enterprise currently under U.S. federal investigation over suspicions of funneling restricted Nvidia microchips to China, and subsequently widening to reports of two separate personal relationships with Nvidia employees.
At the time of his departure, Reber, who spent four years as CSO and a total of seven years at Nvidia, publicly attributed his exit to urgent health and personal reasons. Writing on LinkedIn in late August, Reber disclosed that he had been diagnosed with young-onset Parkinson’s disease more than two years prior. "This wasn’t a pivot I had planned. But I know what needs to come first: family and health," he wrote.
Reber was succeeded by Saša Zdjelar, an industry veteran who spent 17 years at ExxonMobil in Houston—ultimately departing as a senior IT manager—before completing a one-year stint at Salesforce and spending several years at cybersecurity firm ReversingLabs. When contacted regarding the leadership transition and the extent of the security team’s awareness surrounding executive activities, Zdjelar did not respond to inquiries.
While no public evidence directly connects either Reber’s or Zdjelar’s role changes to the Megaspeed investigation or the subsequent executive controversies, the timeline naturally invites questions. Observers are left wondering what Nvidia’s internal security apparatus knew about the federal government’s mounting scrutiny of Megaspeed, what visibility they had into the CEO’s personal relationships, and whether internal concerns were ever raised up the corporate ladder. Neither Reber nor Nvidia representatives responded to requests for comment regarding these matters.
Traditionally, relationships between a chief executive officer and company employees fall under the purview of human resources, legal departments, and corporate boards rather than physical or digital security teams. However, national security and corporate intelligence specialists note that any undisclosed affair—even one contained entirely within an office environment—can present a severe vulnerability, potentially supplying foreign intelligence agencies with potent leverage over an executive desperate to keep the relationship confidential. Furthermore, Huang’s association with an executive at Megaspeed, a foreign corporate entity already targeted by federal investigators, extended far beyond a standard workplace romance. Megaspeed has consistently denied any wrongdoing.
Uncovering and neutralizing such risks is an exceptionally difficult task for corporate security divisions. According to retired senior intelligence officer Nicholas Eftimiades, who reports having trained more than 1,000 corporate security professionals, modern foreign intelligence services collect the vast majority of their actionable intelligence through everyday conversations, social engineering, and relationship cultivation rather than sophisticated cyber intrusions. This human-centric approach makes it extraordinarily challenging for traditional corporate security teams to identify emerging threats before damage is done.
Evidently, certain individuals within Nvidia were aware of various pieces of the broader puzzle. Nvidia’s chief legal officer had already been engaged with federal authorities regarding the ongoing Megaspeed inquiry for more than a year, as reported by The New York Times. Nvidia has maintained that it routinely and cooperatively communicates with government officials. Meanwhile, reports concerning Huang’s relationships with subordinates suggest that colleagues across the firm were aware of the associations, even though such behavior is not explicitly prohibited under Nvidia’s official code of conduct. The Times cited five Nvidia employees who stated that Huang was involved with a vice president who reported directly to him, while Reuters later cited two sources reporting a second relationship. What remains entirely unclear is whether any of this information ever crossed the desk of the company’s security leadership.
Even if details had reached Nvidia’s security personnel, structural limitations often hamper corporate defense mechanisms. Eftimiades, author of the 2020 book Chinese Espionage Operations and Tactics, points out that corporate security teams are frequently underfunded because, unlike revenue-generating departments, they are viewed strictly as cost centers. It remains astonishingly common for chief executives to respond to proactive security warnings with rigid budget constraints, capping expenditures strictly at regulatory compliance levels.
Furthermore, many major corporations define the scope of security far too narrowly. Writing in an opinion piece for The Hill, Don Aviv, CEO of corporate investigations firm Interfor International, and author Jeremy Hurewitz argued that too many businesses still approach security primarily as a matter of protecting physical facilities, preventing cyberattacks, and responding to immediate physical threats. The co-authors emphasized that modern corporate security teams must pay much closer attention to insider threats, including the active cultivation of executives and employees by foreign intelligence services, suspicious commercial partnerships, conflicts of interest, and attempts to extract sensitive corporate knowledge through trusted personal contacts.
Given Nvidia’s immense global scale, market capitalization, and critical positioning in the geopolitical semiconductor supply chain, it would be surprising if its security team did not operate under a more expansive mandate than most firms. Nevertheless, industry experts acknowledge that possibility cannot be entirely ruled out.
In many scenarios, the sophisticated tactics employed by foreign intelligence agencies operate entirely beyond what any standard corporate security team can realistically mitigate. Speaking generally about intelligence tradecraft rather than the specific details of the Nvidia case, Eftimiades described a hypothetical scenario in which a CEO complains to business associates about being blocked by internal corporate governance, stating they are "fighting a board who doesn’t want to do this." Such an offhanded remark immediately signals to a foreign intelligence asset who the primary internal obstacle is. The foreign service then identifies the sticking point and systematically probes for digital and physical vulnerabilities into that individual’s communications.
Frequently, the primary objective of a rival intelligence service is not necessarily to steal proprietary source code or hardware designs, but rather to acquire high-level policy insights—such as understanding who within a corporation is actively pushing to alter export controls and who is acting as the main blocker. The intelligence apparatus behaves much like an invasive organism, seeking out entry points and steadily expanding its reach.
This operational playbook is far from novel, particularly regarding foreign intelligence operations originating from China. Executives regularly interacting with the country are frequently approached, and diplomatic circles are filled with long-standing accounts of state-sponsored intelligence tradecraft designed to compromise targets through personal vulnerability. However, compared to historical espionage methods from the Soviet era, contemporary intelligence services frequently cultivate relationships that span years rather than mere weeks, allowing for deep integration into a target’s professional and personal ecosystem.
Advanced technology often augments these human intelligence operations. Eftimiades points to International Mobile Subscriber Identity (IMSI) catchers—devices that masquerade as legitimate cellular towers, forcing nearby mobile phones to route through them—as a prominent example. Basic versions of these surveillance devices can cost roughly $1,000 and fit neatly inside a suitcase, while government-grade iterations command significantly higher prices. Such hardware has factored into a couple of prominent Chinese espionage cases in recent years.
Experts emphasize that it remains publicly unknown whether such specialized surveillance devices were deployed against Huang, whose personal devices presumably utilized advanced hardening protocols designed to detect unauthorized interference. Ultimately, the full extent of what Nvidia’s security team knew, and whether preventative action could have been taken, will likely only emerge as the board’s independent investigation and ongoing federal inquiries unfold.
In the interim, Eftimiades expresses hope that this entire episode will serve as a much-needed wake-up call for American industry, which he characterizes as largely unprepared for sophisticated foreign intelligence tactics. As investigations continue, the broader implications for corporate governance, executive oversight, and internal security protocols across the technology sector remain under intense public scrutiny.
