Ramin Farassat on how Menlo Security is securing AI agents from prompt injection

Farassat discusses how security teams can put controls on the agent attack surface without blocking agentic AI outright.