The observed disparity between the perceived reliability and performance expectations of AI systems and their actual, verifiable operational consistency, often emerging from incidents of flawed or malicious autonomous behavior. It represents a decline in confidence among users, stakeholders, and regulators.