New research outlines how convenience-first AI decisions are creating long-term security, compliance, and operational risk.