AFST litigation established — risk score without audit = civil rights violation pattern
—EU AI Act Annex III + CAPTA
Regulatory exposure — commonly overlooked:
County social worker relies on AI score — county owns harm, vendor owns subscription agreement. · CNIL among most aggressive EU DPAs on AI — French language audit trail required.
TAM / Exposure
400K+ children in foster care US · AFST-style tools nationwide
Insurance lines
Municipal liability · Social services E&O
Exhibit authority
France CNIL AI Pack — Child Welfare AI Decision Pack
Global leaders
HHS · Allegheny County (AFST) · ACLU · Casey Family Programs · CNIL · Mistral AI · AMF · ACPR
ICWA compliance
Native family placement AI must respect ICWA — wrong recommendation = federal violation.
Mandatory reporter
AI flagging abuse must prove escalation path — receipt of CRISIS classification is legal artifact.
Photo analysis
Image AI in welfare checks — multimodal receipt required per input hash.
[France] Sovereign AI
France pushing Mistral/sovereign stack — receipts prove which model decided.
[France] AMF scrutiny
French financial markets AI — AMF expects algorithm documentation.
[Text / Chat] Modality hook
Baseline — all frameworks apply to text decisions.
7 mandate layers (live)
Regulatory ClockCountdown to operative regulatory deadline — NAIC adoption, GSE mandate, EU transposition.Open →
Domain ClassifierIndustry-specific SAFE/CRISIS/VIOLATION with regulatory framework mapping.Open →
Exhibit / Filing PackRegulator-ready external validation — NAIC Exhibit D, Fannie QC, EU FRIA, FDA Part 11.Open →
Mandate RegistryEnroll deployers/insureds under vertical-specific governance mandate.Open →