NVIDIA just launched its Open Agent Safety Platform

Basically infrastructure to put boundaries around AI agents, trace what they are doing and stop them when they go outside those limits.

Another signal that AI agent security is quickly becoming its own infrastructure layer. But this is mostly from the side of the companies building with AI agents. What happens with the huge majority of fintechs and traditional companies? They are not only going to use AI agents, they are also going to receive attacks from them.

How are they going to defend against AI?

I'm building Inwom(https://inwom.com) for resolving this - Security infrastructure against AI agent attackas