Imagine a future in which a company’s most sensitive exposure is not a stolen password or a breached server, but a private conversation. A relationship, a business introduction or a seemingly ordinary dinner could offer access to people who shape export controls, supply chains and strategic technology.
That possibility sits behind questions now surrounding Nvidia’s security leadership, former chief executive Jensen Huang, and Megaspeed International, a Singapore data-center company linked to scrutiny over the movement of Nvidia chips toward China.
A leadership change under scrutiny
Nvidia identifies Saša Zdjelar as its chief security officer and head of product security in its author page for Zdjelar. The company describes his responsibilities as including corporate and product security, and lists previous experience at Salesforce, ExxonMobil and Crosspoint Capital.
Zdjelar succeeded David Reber, who served as Nvidia’s chief security officer for four years and worked at the company for seven. In his announcement of his departure from Nvidia, Reber said he was leaving for family and health reasons, disclosed his Parkinson’s diagnosis and announced Zdjelar as his successor.
The timing has drawn attention because, about two and a half months later, Huang left Nvidia amid reports concerning relationships with employees and an executive connected to Megaspeed. The reports have not established that the security transition was connected to Huang’s personal life or to the Megaspeed investigation.
TechCrunch reported that Nvidia has not answered questions about what its corporate security function knew or could have done as these issues developed. The publication also emphasized that there is no public evidence tying Reber’s departure to the Megaspeed matter or to Huang’s relationships.
That distinction matters. A change in security leadership is not itself evidence of a security failure. But for a company whose products help power advanced AI systems, the boundary between corporate conduct and national security is becoming harder to see.
The critical case
The strongest criticism comes from lawmakers who argue that the questions deserve federal attention, not only internal review.
In a letter requesting preservation of records concerning Nvidia, Senators Elizabeth Warren and Chris Coons expressed concern about reported connections among Nvidia, Megaspeed, suspected chip diversion and possible intelligence activity. They asked federal agencies to preserve relevant records.
Their argument is not that the public record has already proved espionage. It is that potential relationships involving a strategically important company could affect export controls, access to sensitive business information and the movement of restricted technology. Preserving records, in that view, is a basic step before evidence disappears or investigations become impossible to reconstruct.
Singapore’s Ministry of Home Affairs offers a more limited account. In its statement on investigations into Alice Huang and Megaspeed, the ministry said police and customs began an independent investigation in February 2025 into possible false declarations involving servers. It said the investigation remains ongoing and that Huang was not investigated for espionage.
A wider definition of security
The dispute points to a broader problem for AI companies. Traditional security teams often focus on cyberattacks, physical protection and compliance. Specialists cited by TechCrunch argue that executive conflicts, insider risks, suspicious commercial ties and foreign intelligence cultivation also belong in the security conversation.
Retired intelligence officer Nicholas Eftimiades told TechCrunch that intelligence services often depend on ordinary conversations and long-running personal relationships rather than dramatic hacking operations. The resulting harm might be indirect, such as insight into policy decisions or access to influential executives.
Boards and legal departments may therefore need clearer processes for reviewing executive exposure without turning private life into a surveillance project. Nvidia’s investigations may eventually show a concrete failure, or they may reveal only how difficult influence operations are to identify before they become visible. Either outcome would leave AI companies facing the same question: who is responsible for securing the human relationships around the technology?
- Coolcaesar · CC BY-SA 4.0
This article was generated using AI and published automatically without human pre-publication review.
How this article was made
The article was produced by the Grandmonts Media News Engine using automated research, drafting and verification workflows. No human editor reviewed the article before publication. Grandmonts Media remains responsible for the published content. Errors can be reported at office@grandmonts.cz.