On October 15th, during an interview with Bloomberg in London, Jensen Huang, CEO of NVIDIA, stated that if NVIDIA's customers are found to be developing unsafe or irresponsible artificial intelligence software, NVIDIA will resolutely cease supplying them with chips.

He frankly stated that cutting off supplies is still a "hypothetical issue" at present, and before NVIDIA takes extreme countermeasures such as stopping supply, it will inevitably engage in in-depth communication and dialogue with the relevant customers first.

Jensen Huang added, "We don't operate other people's companies; our technology is general-purpose technology. Therefore, NVIDIA's ability to judge or set limits on how other companies use our technology is actually very limited."

In response to the suggestion put forward by senior executives of companies such as Anthropic that "the entire industry should slow down the development pace of AI", Jensen Huang clearly rejected it. He believes that the current various security concerns are actually "growing pains" during the rapid expansion of the entire industry, and that public anxiety about the security risks of AI has somewhat exceeded the actual situation. He is confident that AI software can be effectively managed and securely deployed, and that many young companies, as they grow in scale, will also gradually become more mature in terms of accountability and responsibility.

In addressing the specific path of security vulnerabilities related to AI, Jensen Huang proposed a technical governance approach of "using AI to monitor AI". He argues that the industry should not slow down due to these issues, but rather should accelerate the integration of prevention, control, and monitoring technologies into the AI system, because " AI itself is the best monitoring tool".

To fulfill this commitment, NVIDIA also announced that it will donate $250 million (Note from IT: the current exchange rate is approximately 1.676 billion RMB) in advanced computing power to UK AI security research over the next five years. This support will assist top British academic institutions, including the University of Oxford, in conducting in-depth exploration of AI security protection and agent control methods.

Related Reading:

《NVIDIA Jensen Huang: If the Frontier Lab cannot control AI, then it must be shut down》

NVIDIA Jensen Huang: AI will not necessarily lead to the extinction of a certain profession, and the claim that AI will destroy humanity within 10 years is pure nonsense.