An AI-skeptical EHS team, building the capability themselves.

The AFL team had seen the AI sales pitch a dozen times. They wanted practical AI that respected the work: incident analysis, procedure review, audit prep.

Larry Pearlman, VP of Environment, Health, Safety and Sustainability (EHSS), wanted two things: a team that uses AI day to day, and a pilot testing whether organizational variables (training cadence, attendance, maintenance overdues) explain injury patterns that lagging indicators can't. His framing: the sites should do this for themselves.

Four cohorts built around the WISER Framework, each taught in its first language: Monterrey in Spanish, Canada and Duncan in English.

Alongside it, a working group across HR, EHS, maintenance, operations, and IT: six hypotheses, plus a narrative workstream that gives operators and supervisors a voice, not only data.

Already true: an EHS function that moved past AI skepticism without losing it, and a shared language for how organizational variables shape safety.

The pilot is running now. The capability program is mid-delivery. Specific outcomes will be reported here when they are earned, not before.