Recognition Before Authority

What AI Should and Should Not Do in the First Minutes of a Working Incident

This position paper by HazNetix co-founder Dan Deppert examines appropriate uses of AI during the first minutes of a working fire or hazardous incident. Written for fire chiefs, command officers, engineers, and public-safety technology leaders, it cites 156 decision points studied among 26 fireground commanders.

Readers will learn how the Display, Recommend, and Act framework separates information support from operational authority, and how to evaluate AI tools against measurable pilot criteria. The cited research found that fewer than 12 percent of decisions involved comparing multiple options, while more than 80 percent reflected recognition of a familiar situation and action. The HazNetix position paper states, “Feed recognition. Do not replace judgment.” It emphasizes visible uncertainty, human control of strategy, and validation before expanded authority.