Nvidia CEO Jensen Huang stated that the company reserves the right to halt chip shipments to clients developing unsafe or irresponsible AI software. While currently a hypothetical scenario, Huang noted that Nvidia would engage with customers before taking such drastic measures. He acknowledged the difficulty in monitoring the use of general-purpose AI technologies across the industry.
Huang rejected calls from industry leaders, such as Anthropic CEO Dario Amodei, to slow down AI development. Instead, he advocates for accelerating the implementation of safety mechanisms. He suggests using AI itself to monitor other systems and manage risks, treating security challenges as a natural part of the industry's evolution.
Nvidia also announced a commitment to invest $250 million in UK computing infrastructure over the next five years. This initiative will support research at the University of Oxford and University College London, focusing on AI safety and the management of intelligent agents to ensure secure technological progress.
