It was four days between the admission of an extramarital affair and the resignation of Jensen Huang, until Thursday the chief executive of Nvidia, the microchip manufacturer at the center of the artificial intelligence boom. The implications of the case, in which the CEO may have handed over key information about his business and industry to the alleged Chinese spy Alice Huang, reflect the interest of Washington and Beijing in being the country that "wins the AI race," putting their intelligence services in pursuit of that goal. In an interview with La Tercera, Dr. Fazl Barez, an expert in AI security and governance at the University of Oxford, discusses the scale of the scandal and its true implications.
What place does Nvidia, the designer of high-end chips, occupy in the rise of artificial intelligence, given that the American company is currently one of the most valuable in the world?
I think the fundamental point is that AI is, today, a technology that is shaping national economies and countries' gross domestic product. Nvidia has become a central piece of this process, since much of the progress in AI depends on access to advanced computing infrastructure. Nvidia occupies a key strategic position, a "chokepoint," given that its graphics processing units (GPUs) and its software underpin much of the current AI capabilities in the United States and, probably, in China as well. This makes it an actor with infrastructure-level power, and not just a successful company. Control over computing capacity determines who can train and deploy advanced models. What makes this especially interesting is that AI is no longer merely a technological development. Its importance is growing for geopolitical reasons, which makes protecting not only confidential secrets but also technological advances far more relevant.
Apple dethrones Nvidia and becomes the most valuable company in the world. What has been the importance of Jensen Huang, and what could change after the resignation of Nvidia's CEO?
I believe Jensen Huang has played a fundamental role in Nvidia's development, especially by placing the company at the center of the current AI boom. However, the degree to which his departure would prove significant or damaging will depend on what changes, if any, occur after his departure. Jensen Huang has been a key figure as the company's leader for a long time, so his departure will undoubtedly have consequences. Although I am not an economist, if the company's value were to fall significantly, this would affect the GDP of the United States. In a broader sense, succession matters because strategic decisions about exports and alliances influence national trajectories, and not just the value of the company itself.
At what point in the "AI race" are we currently, considering the past year, in the contest between China and the United States?
It is not entirely clear exactly where we stand in the AI race. Nevertheless, there is a widespread perception that the United States still holds a slight lead. Even so, I believe Chinese talent and its capacity to catch up at a dizzying pace should not be underestimated. In my view, the real question lies in what this race means and how this competition affects the future of AI and of humanity. I am concerned that we are immersed in a race between the United States and China to develop ever more capable systems, at an extremely accelerated pace. If we want to truly understand how these models work, it may be necessary to slow down. However, this generates pressure in the opposite direction, which discourages any slowdown. It is precisely there that governance collides with competition.
A Chinese AI approaches Claude Mythos in cyberattack tests. To what extent could the information that the alleged Chinese spy Alice Huang may have obtained from Nvidia's CEO be dangerous for the national security of the United States?
As for the alleged incident, the actual scope of the information that may have been compromised is unknown. When the facts are unclear, making exaggerated claims can, in itself, become a security risk. What we can say, with caution, is that supply chains and research ties are part of national strategy. Therefore, transparency and oversight are fundamental. The gravity or harmfulness of this matter will depend on what information was shared. Consequently, protecting not only confidential secrets but also technological advances, for geopolitical reasons, takes on even greater importance.
Fazl Barez, expert in AI security and governance at the University of Oxford.
How close or far is China's AI industry from reaching the American level?
Although it is believed that the United States is ahead, I believe Chinese talent and its capacity to catch up at an astonishing speed should not be underestimated. Chinese researchers should not be disparaged: they are at the same level as their American counterparts. Being "ahead" is not measured by a single indicator, but encompasses factors such as talent, computing capacity, implementation contexts, and the political environment.
Why does Donald Trump say that whoever wins the AI race will win the "race to become the world superpower"? To what extent is this really true?
I have always maintained that AI must be understood in the context of power. International coordination is important, but given the current geopolitical instability, it is very unlikely to be sufficient on its own. In my work on AI-enabled authoritarianism, I argue that democratic accountability can be undermined through surveillance and the use of opaque decision-making systems. Likewise, in the field of interpretability, I observe that technical knowledge does not automatically translate into accountability. Therefore, if democracy is to withstand these pressures, countries also need internal safeguards and transparent institutions.