President Donald Trump shakes hands with Chinese Vice President Han Zhen on the way to his motorcade after arriving at Beijing Capital International Airport in Beijing, China, May 13, 2026. Behind him are Nvidia CEO Jensen Huang and Elon Musk, who were among the technology executives who traveled to China with the president.
Alex Wong | Getty Images
In June 2013, President Barack Obama hosted Chinese President Xi Jinping at his remote California mansion, Sunnylands, for an informal summit. Behind the diplomatic chatter, President Obama has confronted Xi over Chinese state-sponsored cyber espionage and theft of U.S. intellectual property, warning that such activities threaten U.S. businesses and the broader bilateral relationship.
Thirteen years after the Sunnylands Summit, threats have evolved significantly as the race to dominate AI shapes U.S.-China relations. The stakes are higher, the technology is more critical, and the vulnerabilities increasingly include attempts to compromise or co-opt the “key people” at the helm of the nation’s top technology companies.
With the revelation of the scandal involving Nvidia CEO Jensen Huang, the blame should start with the CEO himself, but it shouldn’t end there. This situation illustrates the failure of Silicon Valley, corporate and political leadership on a broader level.
Huang admitted to having an affair with a woman identified in the report as Alice Huang, a former executive at MegaSpeed, a Singapore-based NVIDIA customer. American intelligence officials reportedly believe she was recruited by China’s Ministry of State Security to gather intelligence and potentially influence Nvidia’s efforts to ease American semiconductor export controls. China denies the allegations.
Nvidia’s independent directors have launched an investigation and say they have so far found no leaks of trade secrets or national security information. But the problem goes beyond Huang and Nvidia. Focusing solely on whether proprietary technology or trade secrets were transferred fundamentally misunderstands the potential intelligence value of such operations.
Gaining insight into, influencing, or potentially compromising U.S. technology policy decisions would be equally valuable to the Chinese government. Knowing in advance about export control deliberations, regulatory intent, negotiating positions, and departments within the administration can provide a strategic advantage without stealing a single semiconductor design.
The Trump administration was reportedly alerted to the relationship before the scandal became public. Nevertheless, Huang remained an influential participant in semiconductor policy discussions, advocating greater access to the Chinese market.
The new report also raises questions about communications between Nvidia and Commerce Department officials during the investigation into Megaspeed (which has since been transferred to the Justice Department). While undue influence has not been established, corporate access to government decision-making deserves scrutiny alongside proprietary technology.
A new era of “key man” risk with AI
It remains unclear whether appropriate safeguards were taken. President Donald Trump’s public defense of Hwang, in which Hwang told him he had no idea the woman was a spy, said he trusted the CEO, and said he didn’t remember whether national intelligence agencies had warned him about the threat as reported, conveyed little of the urgency that the allegations of foreign espionage targeting a strategically vital U.S. company warranted.
President Trump has met with Xi three times since returning to power, with two more meetings scheduled for November and December. These talks offer President Trump an opportunity to follow President Obama’s example by directly confronting Mr. Xi over China’s state-sponsored or state-sponsored espionage efforts targeting American technology companies, their executives, and, in some cases, the U.S. policy-making process.
It’s also time for a more formal check from Congress on Silicon Valley, which has long celebrated visionary founders and essential executives whose talents transcend traditional business management and traditional corporate governance. Investors have rewarded this concentration of power, and boards have often responded.
Corporate governance experts refer to this as “key man” risk. This is an organization’s over-reliance on individuals who can cause significant harm to the company through resignation, financial misconduct, or moral compromise. In the age of artificial intelligence, this risk becomes even more dangerous. When a small number of executives control technologies that are central to America’s economic competitiveness, military power, and intelligence superiority, their personal vulnerabilities become potential national security vulnerabilities.
The extraordinary access these executives enjoy to the White House and senior policy makers makes them potential conduits for foreign intelligence gathering, exerting influence not only on their companies but also on the American people and American policy itself.
How can Congress act to take control back from Silicon Valley?
Congress has an opportunity to strengthen America’s defenses and must act now.
The AI Surveillance Act would increase oversight of advanced semiconductor exports and related national security licensing conditions.
The MATCH Act would address weaknesses in restrictions on advanced semiconductor manufacturing equipment, including foreign supplier and service arrangements.
The Chip Security Act would strengthen mechanisms to track sophisticated chips and detect diversion.
Importantly, the Remote Access Security Act extends export control authority to certain remote access arrangements, including cloud computing, and addresses the possibility of accessing limited U.S. technological capabilities without physically exporting the underlying equipment.
Together, these measures can address important technological and regulatory gaps and vulnerabilities. But controlling access to chips, equipment, and computing clearly does not solve the problem of human vulnerability, the risk of so-called “key men” to U.S. technological superiority and national security.
Most of these bills have significant bipartisan support in Congress. However, it encountered fierce resistance from the technology industry and its lobbyists, and very detailed disagreements arose over the details of export control policy. There is also uncertainty about the White House’s support for these bills, with the Republican-controlled Congress not wanting to be on the wrong side of the White House. But there is no doubt in my mind that the NVIDIA controversy has given proponents new grounds for action. If Democrats control both houses of Congress, there will be even more oversight in Congress, and they could be more likely to use these bills to turn many of them into law.
There should also be serious discussion among leading U.S. government and corporate governance experts about whether strategically important technology companies need clear counterintelligence responsibilities at the board level, independent management security assessments, and confidential mechanisms for reporting foreign intelligence activities through appropriate channels.
This does not mean morally policing the private lives of executives. It is about recognizing that the individuals who control America’s most sensitive technologies and influence the policies that govern them are valuable intelligence targets and assets if compromised.
Unfortunately, the Trump administration’s broader philosophy appears to emphasize industry initiative and trust in the ability of technology executives to exercise restraint. Morality is not a governing system. Personal guarantees are not a counterintelligence safeguard. Importantly, shareholder interests are not necessarily synonymous with national interests.
At Sunnylands, President Obama understood that protecting American innovation requires a direct confrontation with China. This engagement helped clarify U.S. red lines against cyber-enabled theft of intellectual property for commercial gain.
President Trump now has the opportunity and responsibility to apply that principle to today’s intelligence threats. Future meetings with President Xi should make clear that state-sponsored or state-sponsored operations targeting U.S. technology executives or attempts to undermine U.S. technology policy decisions are unacceptable and have implications for bilateral relations.
But confronting the Chinese government is only half the responsibility. The US government will also need to confront weaknesses in the set of laws governing access to technology and consider US corporate systems and governance.
The United States cannot declare artificial intelligence essential to national security while leaving the protection of its technology, executives, and policy-making processes primarily at the discretion of corporations. Silicon Valley’s greatest vulnerability may not be the technology we are exporting, but the talent and governance systems entrusted to protect it, and the policies that determine who has access to it.
Neither Congress nor the White House would be reluctant to defer that responsibility to Silicon Valley’s own moral code.
—Dewardrick McNeil, Managing Director and Senior Policy Analyst, Longview Global, and CNBC Contributor. He worked for the U.S. Department of Defense during the Obama administration as an expert on security affairs in East Asia and China.
