How do cloud providers use AI-driven security solutions to mitigate resource over-utilization threats? [closed]

How do cloud providers use AI-driven security solutions to mitigate resource over-utilization threats? [closed]

In cloud environments, unauthorized or malicious resource over-utilization—such as cryptomining or bot-driven compute exhaustion—poses a security challenge by impacting both costs and service availability.

I’m particularly interested in understanding how AI-driven security measures are being used to detect and prevent these types of over-utilization threats. Specifically:

1. How are AI or machine learning models applied by cloud providers to distinguish between legitimate spikes in resource usage (e.g., from a viral ad campaign) and malicious overuse?
2. What specific AI-driven tools or algorithms are used to recognize and limit resource abuse related to unauthorized activities like cryptomining\ Cryptojacking?
3. Are there any known examples of advanced AI implementations that cloud providers use to proactively protect against compute\memory over-utilization (i.e. CPU\GPU, memory) threats?

Note: this question is focused on the security applications of AI within cloud infrastructure to address malicious resource usage patterns rather than operational issues like general monitoring or routine usage limits.