Nvidia says it is building open platform for AI agent security
The chipmaker says stronger safeguards are needed as AI agents take on more critical tasks, though it has released few details on the new platform.
What we know
- Nvidia said in a post on X that it is developing the "NVIDIA Open Agent Safety Platform" with unnamed partners across the industry. [1]
- Nvidia said the effort is meant to help people deploy AI agents "with greater confidence." [1]
- Nvidia said AI agents are taking on more critical work and that "the security behind them needs stronger boundaries." [1]
- Nvidia pointed to comments from CEO Jensen Huang, linked in the post, without detailing what Huang said. [1]
- Not yet confirmed: Nvidia's post does not specify which industry partners are involved, when the platform will launch, or what technical features it will include.
Nvidia said it is building a new platform aimed at securing artificial intelligence agents, according to a post on X from the company's official account. [1]
The company said the initiative, called the "NVIDIA Open Agent Safety Platform," is being developed with partners across the industry, though it did not name them in the post. [1]
"AI agents are taking on more critical work. The security behind them needs stronger boundaries," Nvidia wrote. The company said the goal is to help people put agents to work "with greater confidence." [1]
The post included a link described as comments from Nvidia CEO Jensen Huang, but the post itself did not include further detail on Huang's remarks or on how the platform would function. [1]
Nvidia did not say in the post when the platform would be released or which companies are participating as partners.