The Software & Information Industry Association (SIIA) submitted a comprehensive response to the National Institute of Standards and Technology (NIST) regarding its responsibilities under the Executive Order on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.
A key recommendation is the development of a companion resource for generative AI (GenAI), building on the success of NIST’s AI Risk Management Framework (AI RMF). SIIA advises leveraging existing standards like ISO/IEC 42001 and collaborating with industry leaders to address the unique risks associated with GenAI, emphasizing distinctions among actors in the AI value chain.
The submission highlights the necessity of risk management tailored to GenAI’s socio-technical nature, emphasizing the importance of context, language use, and cultural variations. Additionally, SIIA advocates for impact assessments before deployment, ongoing monitoring, and heightened assessment requirements for GenAI tools impacting legal rights or essential services.
Addressing synthetic content risks, SIIA supports NIST’s efforts and proposes technical solutions, such as engaging with DARPA’s Semantic Forensics team, while stressing the need for digital literacy programs. SIIA also encourages NIST’s role in global alignment on AI standards, emphasizing inclusion of smaller enterprises in international standardization efforts. The submission reflects a comprehensive and collaborative approach to advancing responsible AI development and global technical standards.