Model drift (also called concept drift) occurs when the data used for training no longer reflects the current environment. In marketing, consumer behaviors, trends, and demographics change rapidly; therefore, a model trained on five-year-old data will suffer from significant drift. The AAIA™ manual explains that this results in the model making predictions based on obsolete patterns, leading to poor accuracy and potentially biased or irrelevant marketing recommendations. Options A, B, and C refer to security attacks, whereas " Model drift " is the natural operational degradation that occurs when AI systems are not regularly updated with fresh, representative data.