News · August 18, 2022
NIST's second AI risk framework draft puts governance first
The US standards body published a second draft of its voluntary AI Risk Management Framework. It is not regulation, but it is likely to become the checklist customers and auditors ask about.
What happened
- NIST said on August 18, 2022 that it was seeking comments on the second draft of its AI Risk Management Framework, which is intended for voluntary use, with comments due September 29 and a workshop on October 18-19.
- According to NIST, the framework is built around four functions, Govern, Map, Measure and Manage, and a draft online Playbook with suggested actions currently covers Govern and Map; NIST planned to publish version 1.0 in January 2023.
- Davis Wright Tremaine reported that the draft lists seven characteristics of trustworthy AI, among them validity and reliability, safety, fairness with bias managed, security, transparency, explainability and privacy, and moves Govern to the first position as a cross-cutting function.
- The firm also noted the draft's emphasis on testing, evaluation, verification and validation across the AI lifecycle, and predicted it could shape industry standards much as NIST's cybersecurity guidance has.
Why it matters
Voluntary frameworks have a way of becoming expected. If you sell software that uses machine learning to larger companies, procurement questionnaires will likely start borrowing NIST's language, so it pays to know it before the questions arrive.
A light first step: name an owner for AI risk, keep a short inventory of where models are used, document how each was tested and on what data, and decide when a human reviews outputs. Small teams can do this in a few pages, and it makes later reviews much easier.
Sources
- NIST: NIST Seeks Comments on AI Risk Management Framework Guidance, Workshop Date Set
- Davis Wright Tremaine: NIST Releases Second Draft of AI Risk Management Framework
This post is CMM’s own summary and commentary on publicly reported news. Facts are attributed to the sources linked above; please read them for full details. CMM is not affiliated with these publishers or with the companies mentioned, and trademarks belong to their owners. Prepared with AI assistance. Corrections or requests: [email protected]. See our editorial policy.
Want help with this?
Tell us what you need, and we will suggest the right team model on a short call.
Talk to us