Dear Kenneth,

AI is often positioned as a way to increase efficiency — but in regulatory affairs, reliability matters just as much as speed.

This study compares six leading AI models against human expert performance across common regulatory tasks.

The results highlight a clear pattern: while some models perform well in specific scenarios, no model consistently delivers the accuracy and completeness required for independent use.

For regulatory, quality, and compliance leaders, this raises important considerations about governance, risk, and practical adoption.