What is Human Oversight of AI?

Definition

Human oversight means a qualified person can monitor, intervene, or stop an AI system when risk spikes. It is broader than a single appeal right because it covers ongoing monitoring, not only customer complaints.

See also our AI compliance glossary for short definitions of common terms.

Which regulations require this

• Washington SB 5395 - AI in Health Insurance Prior Authorization

  Regulates use of AI in health insurance prior authorization decisions. Requires human review of AI-generated denials.

  WAAI-specificMediumUpcoming
• Connecticut SB 1103 - Automated Employment Decision Tools

  Requires employers using automated decision tools in employment to conduct impact assessments, provide notices, and a...

  CTAI-specificMediumIn effect

Which states reference this obligation

What you should do next

• Name accountable owners for each production model with clear escalation paths.
• Define when automation must pause, for example abnormal error rates or drift alerts.
• Train operators on what they can change without a full redeployment cycle.
• Log overrides and the business reason so you can show proportionate judgment.
• Test oversight playbooks during tabletop exercises at least annually.