Jensen Huang: Cybersecurity Is the Next Big Market for AI Computing Power

Jensen Huang said plainly that AI's automation of software development comes at a cost - code written and patched at machine speed opens new avenues for attack. Nvidia's CEO framed this as a business opportunity, not just a threat, and said cybersecurity will become the next major driver of demand for his company's computing power.

What Huang Said

Speaking at the Goldman Sachs technology conference in San Francisco, Huang said cybersecurity will likely be the next major application of artificial intelligence. In his view, AI systems capable of writing code should also be able to find vulnerabilities in it and patch them, and that requires enormous computing resources.

What's better for creating demand than creating a problem. Who wouldn't want the market to go crazy for their product and line up for it - Jensen Huang, Nvidia CEO

The remark, though delivered with a wink, reflects a real corporate strategy. Huang acknowledged that AI progress simultaneously makes systems more vulnerable to attack and fuels demand for his own company's products that defend against those attacks.

Attacks at Inference Speed

Huang warned that attacks on companies will grow exponentially, since autonomous AI systems let criminals automate assaults at a scale never seen before. He said traditional security metrics, like breakout time, are losing their meaning as attacks start happening at the pace models operate, that is, at inference speed.

A week earlier, on September 1, Nvidia and CrowdStrike presented the SafeMind system at the Fal.Con conference in Las Vegas, described as the cybersecurity industry's first comprehensive agentic system. It runs on two models: a defensive one called Blue Solano and an offensive one called Red Tempest, which compete against each other in a digital twin, learning from each other's simulated attacks.

Partnerships and Numbers

Nvidia's Nemotron models are set to be fine-tuned on CrowdStrike's security data, creating threat-detection systems tailored to enterprise environments. Huang also cited partnerships with Cisco and Palantir as part of a broader strategy to position cybersecurity as a third pillar of Nvidia's revenue growth, alongside data centers and AI infrastructure.

Market reaction was mixed - CrowdStrike shares fell nearly 7 percent after the partnership was announced, while Nvidia's dropped about 1 percent, amid broader market weakness. Huang nonetheless reaffirmed his forecast of 70 percent revenue growth for the company in the coming fiscal year, noting that unmet demand for its products exceeds even 100 percent.

What It Means for Companies and Investors

For companies, this means the market for AI-based defensive tools will grow in step with the market for threats that the same technology generates. Cloud and cybersecurity providers in Poland can expect growing pressure to deploy real-time attack-detection systems, since according to Nvidia, classic response windows measured in minutes are no longer enough.

For investors, Huang's remarks are another signal that Nvidia is looking for new justifications for continued, massive spending on AI infrastructure, at a time when forecasts point to as much as $3-4 trillion in global outlays by 2030. Critics of this narrative note that the company has a direct interest in highlighting every new AI application that drives sales of its chips.

Huang also suggested that the model of competing defensive and offensive agents, developed for cybersecurity, could eventually be applied to robotics, industrial manufacturing, and edge systems, fitting into Nvidia's broader strategy of building complete 'AI factories' rather than just graphics chips.