The Center for Security and Emerging Technology (CSET) offered this submission for NIST’s consideration in response to its draft Standards for Reliable, Robust, and Trustworthy Artificial Intelligence. NIST requested information about “the current state, plans, challenges, and opportunities regarding the development and availability of AI technical standards and related tools, as well as priority areas for federal involvement in AI standards-related activities” and listed 18 topics of interest. CSET’s submission is on the following topics:
- Topic 3: The needs for AI technical standards and related tools. How those needs should be determined, and challenges in identifying and developing those standards and tools.
- Topic 11: Specific opportunities for, and challenges to, U.S. effectiveness and leadership in standardization related to AI technologies.
CSET’s submission focuses specifically on standards for “reliable, robust, and trustworthy” AI (collectively “trustworthy AI”), as that phrase is used in the Executive Order on Maintaining American Leadership in Artificial Intelligence. It expands on the following claims:
- Standards for trustworthy AI can advance both U.S. and global interests. Standards are a prerequisite for intra- and international coordination toward trustworthy AI.
- Developing standards for trustworthy AI requires foundational research. Standards should be informed by foundational research on what makes AI trustworthy.
- NIST should create a National AI Testbed. NIST should advance foundational research on trustworthy AI by establishing a National AI Testbed: a digital platform containing public and non-public datasets, code, and testing environments on which AI systems can be developed, stored, and tested.