NIST PRINCIPLES FOR EXPLAINABLE AI

Start date :2020
End date :
Name of responsible governmental body :
  • National Institute of Standards and Technology (NIST;)
Description :
  • The National Institute of Standards and Technology (NIST) developed a draft report intended to stimulate a conversation about what NIST should expect of decision-making devices, as part of a broader NIST effort to help develop trustworthy AI systems.
United States flag
Objectives:
  • To provide guidance that AI systems should deliver accompanying evidence or reasons for all their outputs.
  • To provide guidance that AI systems should provide explanations that are meaningful or understandable to individual users.
  • To provide guidance that such explanations correctly reflect the system's process for generating its output.
  • To provide guidance that AI systems should only operate under conditions for which they were designed or when they reach a sufficient confidence in their output.
OECD AI principle adressed:
  • Transparency and explainability
  • Accountability
Relevant policy area covered:
  • Agriculture
  • Competition
  • Development
  • Digital economy
  • Economy
  • Education
  • Employment
  • Environment
  • Finance and insurance
  • Health
  • Innovation
  • Investment
  • Public governance
  • Science and technology
  • Social and welfare issues
Related to a theme:
Direct beneficiaries:
Estimated budget expenditure (range per year):
  • Don't know
Background including shifts in the policy initiative:
  • N.A
Policy initiative is a structural reform:
  • No
Evaluated:
  • No
Link to evaluation or reports:
  • N.A
Public access URL:
Funding from the private sector:
  • No
Type of evaluation:
  • N.A
Evaluation provides input to:
  • N.A
Evaluation performed by:
  • N.A
Policy instruments :

Regulatory oversight and ethical advice bodies

  • Type : Regulatory oversight and ethical advice bodies
  • Name in English : National Institute of Standards and Technology
  • Country : United States
  • Public access URL : https://www.nist.gov/publications/four-principles-explainable-artificial-intelligence-draft
  • Type(s) of oversight or advice : Technical guidance (e.g. toolkits, documentation, technical standards)
  • Activities : Gather opinions from stakeholders on ethical principles, regulation improvements, etc.
  • Reports to : Ministry
  • The regulatory oversight/ethical advice body is composed of : Mostly government representatives
  • Reports are publicly available : Yes

Emerging AI-related regulation

  • Type : Emerging AI-related regulation
  • Name in English : National Institute of Standards and Technology Principles for Explainable AI
  • Country : United States
  • Public access URL : https://www.nist.gov/system/files/documents/2020/08/17/NIST%20Explainable%20AI%20Draft%20NISTIR8312%20%281%29.pdf
  • Role of government : Technology/innovation enabler (e.g. interoperability standards)
  • Challenge(s) addressed : Transparency and explainability
  • Type(s) of regulation : Self-regulation (e.g. codes of conduct, guidelines, standards)
  • Regulatory approach : Technology-based regulation (e.g. moratoria, bans, standards of use)
  • Level of governance : National
  • Approach to monitor compliance : Regulated parties are incentivised to adopt monitoring technology that is not managed by the regulator
Last updated : Sep 21, 2022
OECD and European Commission Logo