Return research task list
M-07In ProgressNorth America / EuropeFinance

Risk Management and Regulatory Compliance in Financial AI

Credit, anti-fraud, and market risk models improve efficiency but also introduce model risk into decision chains. This task reviews financial institutions' model governance practices and regulatory responses, documenting known failure scenarios.

Target Case
30row
Submitted
15row
Deadline
2027-01
Completion progress
50%

Research Methodology

The method determines the form of evidence, and the form of evidence determines how far the conclusion can go.

  • Model Governance Document Analysis
  • Compliance Process Interview
  • Post-mortem of failed cases

Delivery Requirements

  • 01Model Governance Process Overview
  • 02Compliance Requirements Comparison Table
  • 03Invalidation and Rollback Records

Acceptance Criteria

  • 01Specify model type and decision-making authority.
  • 02Provide traceable evaluation metrics
  • 03Must include specific regulatory compliance requirements

Standard of Evidence

Consistent with A-AO research standards.

  • Each case must have at least 2 pieces of evidence, including at least 1 that are either primary sources or authoritative secondary sources.
  • Resales (including materials provided by enterprises) cannot serve as the sole basis for support and may only be used for comparison.
  • Sources must be traceable: values without sources are excluded from the case body.
  • Evidence levels and case verification levels are recorded separately and cannot be converted into one another.

Task Owner

Professor Hans Müller · ETH Zurich

Method Design and Acceptance Criteria for This Task

View Researcher Directory →