The news: Jensen Huang resigned as CEO of Nvidia after admitting to having an extramarital affair with a woman who US officials suspect is a Chinese spy. Huang co-founded Nvidia in 1993 and is seen as being largely responsible for its becoming one of the most valuable companies in the world as lead chipmaker for the AI boom.

Huang’s affair accelerated into a national security concern given US chip export controls on China, which is considered a top AI rival that would have an interest in acquiring sensitive information about Nvidia.

Why this matters: The stakes behind Huang’s resignation reveal the murky waters that businesses using AI are swimming in. On the surface, AI is a tool for powering productivity, creative production, or novel customer experiences, but in the depths it’s a swelling geopolitical powder keg. Disentangling these two sides of the technology is impossible.

The concentration of power in the AI industry adds even more instability for enterprise users. Nvidia commands 81% market share of AI chips by revenues, per IDC data cited by Yahoo, supplying virtually every AI player that matters, including OpenAI, Google, Anthropic, Meta, Amazon, and Microsoft. Many of these companies—which are enterprise partners themselves—have their own deals with each other, creating a nested circularity that’s at risk of unraveling if just one major component slips.

Businesses shouldn’t consider themselves safe just because they’re free from dealing with AI hardware directly. Those that use the technology are downstream from any potential fallout if, for example, Nvidia’s chip production falters and leads to a cascading shortage across the data centers AI relies on.

Implications for businesses: Giving up AI entirely isn’t a viable option, but neither is hitching one’s wagon to the boom or bust of such a precarious technology. Businesses need to take this moment to consider their investments in AI and how deeply they want to rely on it going forward, as its geopolitical baggage is likely to only intensify.

Developing strategies for risk management—such as by limiting AI experiments to certain areas of the business or exposure to companies at the center of the race—should be viewed as a core responsibility of AI governance.

This isn’t merely an issue of future uncertainty—today, AI is showing very little payoff for enterprises despite the size of their investments. Making key decisions now could help protect businesses in the long term, and doing so requires a deep dive into the technology’s cost-benefit tradeoff.

This content is part of EMARKETER’s subscription Briefings, where we pair daily updates with data and analysis from forecasts and research reports. Our Briefings prepare you to start your day informed, to provide critical insights in an important meeting, and to understand the context of what’s happening in your industry. Non-clients can click here to get a demo of our full platform and coverage.