Run threat assessments through multiple AI models and log where they disagree....

https://jaspersmasterinsights.opalvector.com/posts/what-is-a-multi-llm-orchestration-tool-and-why-do-teams-use-one

Run threat assessments through multiple AI models and log where they disagree. Write down every assumption, name the person who makes the final call, and set a follow-up review date for each case