RESPONSIBLE AI, COMPLIANCE, AND RISK

CHAPTER SUMMARY Bias can enter through training text, so agents that affect people need matched-pair fairness tests. Privacy practice means minimizing, masking, protecting, and deleting personal data. Auditability needs tamper-evident records kept for the required period. Frameworks such as the NIST AI Risk Management Framework guide risk rating, and human oversight, escalation, and named accountability complete the governance layer. KEY TERMS REVIEW Bias: a systematic unfair tilt in outcomes. Fairness: similar people receiving similar treatment. Matched-pair test: comparing inputs identical except for one detail. Data minimization: collecting only the data you need. Audit: an independent review of what a system did. Tamper-evident: changes would be detectable. Retention period: how long records must be kept. Escalation path: a route for sending hard cases to a human. Accountability: a named party responsible for results.