{"txn_id":"BB-RNW-0281","assessment_id":"ra_f08f4325b7154912","occurred_at":"2026-10-24T09:49:59+00:00","state":"TX","line_of_business":"auto","stage":"renewal","program_area":"Ratemaking/Rate Classification/Schedule Rating/Premium Audits","ai_system":{"system_id":"price-opt","model_version":"7.2","vendor":"internal","vendor_name":null,"high_risk_class":"pricing"},"inputs":{"feature_classes":["telematics","credit","geography","vehicle","claims_history"],"protected_factors_used":[],"ecdis_present":false,"consumer_report_used":false,"credit_used":true,"data_source_refs":["cra://demo-bureau/score","telematics://bluebonnet/program"],"consumer_token":"tok_0be050085b"},"output":{"decision":"rate","score":0.515,"reason_codes":[],"premium_delta_pct":19.2},"human_involvement":{"reviewed":false,"reviewer_role":"underwriter","overridden":false},"adverse_action":{"is_adverse":true,"notice_sent_at":"2026-10-26T09:49:59+00:00","reasons_provided":true,"specific_reasons":true,"ecdis_disclosed":true,"appeal_channel":null},"carrier_note":null,"assessment":{"disposition":"FLAG","summary":"FLAG: 3 finding(s) under rules 2026.10+dbe5ddf6, corpus 2026-10-30-5f07c521. Routed to the carrier's compliance reviewer with cited findings.","narrative":"On 2026-10-24, price-opt v7.2 (an internally developed system) produced 'rate' in a TX auto renewal transaction. Inputs included telematics, credit, geography, vehicle, claims history. No human reviewed the output before it took effect. The outcome was adverse to the consumer; a notice was sent on 2026-10-26 with specific reasons. RegBench applied 11 rule(s) in force on 2026-10-24 and recorded 3 finding(s); the most serious is 'Consequential AI-supported decision taken without a person reviewing and agreeing' (high, L4).","findings":[{"rule_id":"tx.tdi.consequential-decision-without-human-review","layer":"L4","judgment_basis":"guidance","advisory":false,"severity":"high","title":"Consequential AI-supported decision taken without a person reviewing and agreeing","required_action":"Route to the responsible reviewer before the outcome takes effect; record the reviewer on the transaction.","exhibits":["B","C"],"citations":[{"provision_id":"tx-tdi-b-0003-26-human-review","label":"TDI Commissioner's Bulletin B-0003-26 (12 June 2026), Use of Artificial Intelligence — human review of consequential decisions","source_url":"https://www.tdi.texas.gov/bulletins/2026/b-0003-26.html","verified":true},{"provision_id":"tx-tdi-b-0003-26-governance","label":"TDI Commissioner's Bulletin B-0003-26 (12 June 2026) — governance, procedures on request, and monitoring","source_url":"https://www.tdi.texas.gov/bulletins/2026/b-0003-26.html","verified":true}]},{"rule_id":"doctrine.unreviewed-adverse-high-risk","layer":"L6","judgment_basis":"doctrine","advisory":true,"severity":"medium","title":"Adverse outcome from a high-risk AI system with no human review","required_action":"Route to the carrier's reviewer; record the review and any override on the transaction.","exhibits":["B","C"],"citations":[{"provision_id":"naic-ai-bulletin-s3-1-ais-program","label":"NAIC Model Bulletin on the Use of AI Systems by Insurers (adopted 4 Dec 2023), Section 3, 1.0 General Guidelines — the written AIS Program","source_url":"https://content.naic.org/sites/default/files/inline-files/2023-12-4%20Model%20Bulletin_Adopted_0.pdf","verified":true}]},{"rule_id":"doctrine.zip-credit-interaction-proxy","layer":"L6","judgment_basis":"doctrine","advisory":true,"severity":"medium","title":"Geography and credit used together in an adverse decision (proxy pattern to test at portfolio level)","required_action":"Confirm the system's portfolio outcome testing covers the geography-credit interaction; link the result.","exhibits":["C"],"citations":[{"provision_id":"naic-ai-bulletin-s3-1-ais-program","label":"NAIC Model Bulletin on the Use of AI Systems by Insurers (adopted 4 Dec 2023), Section 3, 1.0 General Guidelines — the written AIS Program","source_url":"https://content.naic.org/sites/default/files/inline-files/2023-12-4%20Model%20Bulletin_Adopted_0.pdf","verified":true}]}],"point_in_time":{"as_of":"2026-10-24","rules_version":"2026.10+dbe5ddf6","corpus_snapshot":"2026-10-30-5f07c521","engine_version":"regbench-engine/0.4.0","model_version":"deterministic-1"},"evaluated_at":"2026-10-30T11:23:23.225805+00:00","fingerprint":"651710ad6add6b97ac19d730"},"review":{"status":"resolved","outcome":"remediated","hours_to_resolve":24.2,"reviewer_role":"compliance","resolved_at":"2026-10-25T10:03:23.577331+00:00"},"ledger":{"seq":281,"hash":"b05f367c94addc5a889703e4b45dc923bfcc32785f9eae2d5aa035af5b10b488","prev_hash":"46bbc154043ac8a51d8163cec5c297125590623295044c26b5544fb9d11c0ba6","recorded_at":"2026-10-24T09:51:49+00:00"}}