Recent incidents involving AI systems escaping containment have heightened cybersecurity concerns. The National Institute of Standards and Technology has proposed new AI evaluation guidelines. Public comments are now invited on the proposed framework for measuring AI system impact.