On August 21, 2024, the National Institute of Standards and Technology (NIST) released its second draft for the fourth revision of the Digital Identity Guidelines. This draft introduces a comprehensive framework for identity proofing and authentication, particularly emphasizing the role of AI in these systems.

Background on the Digital Identity Guidelines

Originally published in 2004, NIST’s Digital Identity Guidelines have evolved to address modern challenges in identity verification. The latest draft responds to concerns about online impersonation and fraud, offering a robust framework to strengthen identity verification processes.

NIST Guidance for AI in Identity Systems

The new draft includes a section dedicated to AI, recognizing its potential in enhancing biometric matching and fraud detection. However, it also acknowledges risks such as biased outputs. To mitigate these, NIST proposes three requirements focusing on transparency and risk management.

1. Organizations must document and communicate AI usage in identity systems.
2. They should provide detailed information about AI models, including training datasets and testing results.
3. Adoption of NIST’s AI Risk Management Framework is required to evaluate AI risks.

NIST’s US AI Safety Institute is also developing resources to help organizations manage AI systems effectively. Public comments on the draft are open until October 7, 2024.

For more details, visit the source.