Former regulator urges new approach to AI explainability
Ex-OCC chief Michael Hsu suggests shift from academic analysis to decision-based techniques
Former acting US comptroller of the currency Michael Hsu has called for banks and regulators to rethink how they handle the explainability challenge around large language models. He wants to see them take a more decision-focused approach that exploits the unique nature of LLMs.
Explainability – the ability to follow the logic that leads to the outputs from artificial intelligence – has been one of the biggest hurdles to LLM adoption among banks. By design, LLMs involve a vast number of input
More on Risk management
How AI agents can join the dots for risk managers
Citi risk expert outlines agentic AI tool that would pull together structured and unstructured data on trading and lending approvals to create single, unified view of risk
In Iran war, VAR models ease cliff effect on Ice and CME margins
At 105%, EEX – using Span model – saw largest single-day jump compared with those CCPs
Newcomer of the year: Abaxx Exchange
Energy Risk Awards 2026: New exchange sets out to modernise commodity derivatives by aligning them to physical markets
AI project of the year: SOCAR Türkiye
Energy Risk Awards 2026: Risk team harnesses AI to transform RCSA into a scalable, sustainable and internally owned capability
Data, cyber and model risk top IT concerns for risk managers: survey
Energy Risk software survey reveals risk managers’ tech pain points and plans
Energy Risk Debates: the Iran conflict and the widening mandate of the risk manager
Panellists discuss the impact of the Middle East crisis so far on risk teams and the drive towards enterprise risk management
Abaxx: meeting the need for new commodity derivatives
Abaxx revamps commodity hedging with a suite of modern contracts
Tokenised commodities could help oil the machine
Shifting physical assets onto the blockchain eases collateral frictions, argues crypto expert
