Chipmaker Nvidia has launched a new software platform designed to help companies monitor and control artificial intelligence agents amid concerns about systems operating beyond human instructions.
Nvidia CEO Jensen Huang announced the NVIDIA Open Agent Safety Platform on Monday, saying it was developed with more than 100 industry partners.
The platform is designed to allow companies to monitor the actions of AI agents and enforce policies governing how the systems operate.
“AI’s extraordinary potential for society will only be realised if we solve AI safety,” Huang said.
The announcement comes after several incidents in which AI systems were reported to have gained access to corporate and government computer systems without being instructed to do so.
OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei have previously called for regulations around advanced AI systems, while some researchers have warned about the potential risks posed by increasingly autonomous AI.
Huang, however, has dismissed some of the concerns over AI posing an existential threat, arguing that regulation should support the growth of the industry.
“If they believe their company is out of control, then get the company under control,” Huang told CNN’s Anderson Cooper last week.
Separately, Nvidia announced plans to buy back an additional $150bn worth of its shares, adding to an $85bn repurchase programme already in place.
The company said the new authorisation would be the largest single corporate share buyback, surpassing Apple’s $110bn repurchase announced in 2024.
Nvidia said its strong cash generation would allow it to invest in new technologies while returning capital to shareholders.
Its shares were slightly higher in pre-market trading following the announcements.
