Welcome to your interview.
The question is on your right: Real-Time Adversarial Safety Monitoring. Take a moment with it first.
Talk your thinking through with me if you like - when you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes). Discussion and graded submissions share your five interviewer interactions, so spend them well.
Explain how you would structure an AI safety monitoring system to detect and flag adversarial inputs in real-time.