In response to President Biden’s Executive Order (EO) Concerning Artificial Intelligence (AI), the National Institute of Standards and Technology (NIST) held a Request for Information (RFI) on creating guidance for the development, deployment, and evaluation of AI technologies. In a new insight, Technology and Innovation Policy Analyst Joshua Levine walks through the components of AI policy covered in the RFI, emphasizing the benefits incorporating existing and developing standards from industry, civil society, and multistakeholder groups could have on the guidance’s diffusion and impact.

Levine concludes:

If approached wisely, NIST can lay the foundation for standards that could help promote innovation and competition within the market for AI-powered technologies. By providing guidance for developers and deployers on issues related to model development, red-teaming and stress testing, appropriate data practices, and other issues, NIST can encourage firms and individuals to build models that are safe, secure, and trustworthy. Further, as guidance is developed, looking to the private sector, multi-stakeholder organizations, and other nations can provide useful references in areas such as watermarking and content provenance. To achieve the goal outlined by the AI EO and the RFI, NIST should consider how previous documents and standards it and other bodies have promulgated can contribute to the present and future iterations of guidance.

Read the analysis