Addressing the G20 Innovation Ministerial in North Carolina on Wednesday, NVIDIA's chief executive urged world leaders to embrace artificial intelligence infrastructure, warning that excessive fear or heavy-handed regulation could cause nations to miss out on the next major technological breakthrough.

The CEO made the case that AI infrastructure should be viewed as fundamental to a country's economy, on par with electricity, roads and the internet. He stressed that every nation must build the necessary infrastructure to support its own economic growth.

He argued that domestic AI infrastructure can provide computing power to researchers, students, startups and various industries, helping them develop new products and expand business opportunities.

While acknowledging potential risks, the NVIDIA chief suggested the greatest danger does not actually stem from the technology itself. The worst possible outcome, he warned, is failing to seize the opportunity and being left behind by progress. That, he emphasised, is the only truly devastating result.

He cautioned that countries cannot advance if they approach this technology with fear, uncertainty and anxiety. A balanced approach is essential, he said, noting that governments do not need to compete at every level of the AI technology stack. Instead, they should identify where they hold a competitive advantage while promoting AI adoption across education, healthcare, manufacturing, scientific research and other sectors.

The CEO also highlighted that AI extends beyond software alone. Large language models can evolve into intelligent agents with memory, information retrieval and tool-calling capabilities, ultimately powering robots, autonomous vehicles, manufacturing systems and scientific laboratories.

He pointed to the United States as an example of significant AI progress, driven by favourable energy policies, regulatory conditions and speed of development. He noted that US infrastructure could attract nearly $1 trillion in investment this year alone.

The NVIDIA leader also pushed back on calls for broad restrictions on AI, arguing instead that regulation should target measurable, real-world risks rather than hypothetical scenarios. He advised against regulating assumed theoretical harms, saying the focus should be on actual, tangible dangers that genuinely materialise.

He emphasised that companies developing AI bear the responsibility for ensuring the technology's safety, while regulators should concentrate their efforts on genuine concerns. He drew a parallel with the aviation industry, where airlines manage safety protocols while passengers simply focus on where the technology can take them.

He concluded that AI remains in its early stages, and technological advancements in fact-checking, output accuracy and scientific reliability are already making the technology safer.