
Cowater had been discussing AI for over a year, and had invested in an AI roll-out, but this had not yet translated into practical change at scale. Experimentation was uneven and often private, there was no shared view on the potential for operational value, and there was no clear mechanism for turning individual trials into shared practice. People were curious, but incentives to experiment and share were weak.



