Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Artificial Intelligence topic
No spam. Unsubscribe anytime.
Treasury and MRA outline AI, fraud and cyber workstreams to help smaller institutions
Summary
Treasury and advisory‑committee panels described joint public‑private workstreams to close data and capability gaps in AI, fraud detection, identity, and cyber resilience, with specific initiatives (AI lexicon, explainability, AI 'nutrition labels') to help smaller institutions comply with explainability and operational standards.
Treasury Department officials and MRA panels on Thursday outlined coordinated workstreams to help financial institutions manage risks from artificial intelligence and cyber threats, especially to smaller banks and market participants.
Todd Conlin, Treasury's Chief Artificial Intelligence Officer (Deputy Assistant Secretary), summarized the Treasury AI report and described multiple sector workstreams: creating a common AI lexicon, developing explainability and accountability practices, an AI 'nutrition‑label' construct for third‑party data, and fraud‑data workstreams aimed at narrowing data gaps for smaller institutions. Treasury and the New York Fed will co‑lead taxonomy and definitions; industry partners (PNC, CRI, ABA) will pilot explainability templates and nutrition‑label prototypes.
Panelists said most large financial institutions already embed AI in cyber defenses and fraud detection, but smaller institutions lack data, developer capacity, and procurement access. Treasury reported planned public‑private efforts to produce operational roadmaps, vendor labeling, identity/authentication guidance and training to scale responsible AI use and to preserve competition and security across the sector.
No formal committee vote was taken on the AI workstreams; Treasury framed them as ongoing collaborative projects with deadlines and deliverables through late 2025.
