We need to start with big pictures of threats and impacts related to AI applications
Could we use realtime threat modeling to monitor for new and evolving threats?
Wouldn't this just blur the line between threat modeling and attack/abuse/system modeling?
Real-time threat modeling is an interesting idea, but I’m not sure we’re there yet (assuming a continuous, complete, and automated process).
That said, threat modeling can be effectively integrated with metrics, monitoring, assurance, or penetration testing (and greatly improve from that!)
Could we use realtime threat modeling to monitor for new and evolving threats?
Wouldn't this just blur the line between threat modeling and attack/abuse/system modeling?
Real-time threat modeling is an interesting idea, but I’m not sure we’re there yet (assuming a continuous, complete, and automated process).
That said, threat modeling can be effectively integrated with metrics, monitoring, assurance, or penetration testing (and greatly improve from that!)