Every Frontier LLM Fails at Threat Hunting: Best Model Finds 3.8% of Malicious Events When Given Raw Logs and No Hints

Every Frontier LLM Fails at Threat Hunting: Best Model Finds 3.8% of Malicious Events When Given Raw Logs and No Hints | BedrockNews