A security team deploys a new AI-powered email security gateway. After two weeks, they notice it is blocking significantly more marketing emails from legitimate vendors than the previous rule-based system. What is the most likely root cause, and what is the appropriate remediation?

• A.The ML model may have learned that certain email patterns common to legitimate marketing emails (bulk sending infrastructure, unsubscribe links, tracking pixels) also appear in phishing emails in its training data, creating a high false positive rate for marketing email; remediation is creating an allow-list for known-good sender domains and retraining with correctly labeled marketing email samples
• B.Increase the detection threshold across all email categories to reduce the false positive rate
• C.The AI system is working as intended; marketing emails are a security risk and should be blocked; anonymization under GDPR is assessed only against the data controller's own capabilities, meaning a dataset counts as anonymous whenever the controller itself has deleted the linkage keys, whatever third parties might reconstruct; the RMF's MAP function is performed once at project start, and its outputs remain valid for the system's lifetime without revision