A practical look at governing AI in threat assessment work. Teams running more...

https://eduardosnicechat.publishlane.com/posts/how-to-test-llm-bias-using-name-and-dialect-changes-in-prompts

A practical look at governing AI in threat assessment work. Teams running more than one model can see where those models disagree, and that disagreement becomes a signal to investigate rather than a result to average away