Luján, Lofgren Lead Letter Urging Multilingual Translation of the NIST AI Risk Management Framework

Washington, D.C. – U.S. Senator Ben Ray Luján (D-N.M.), Chair of the Subcommittee on Communications, Media, and Broadband, and U.S. Representative Zoe Lofgren (D-Calif.), Ranking Member of the House Committee on Science, Space, and Technology, respectively, urged the National Institute of Standards and Technology (NIST) to translate the AI risk management framework (AI RMF) into non-English languages. 

The United States is a leader in the development of artificial intelligence systems, and must lead the world in responsible artificial intelligence governance as well. NIST has developed a strong foundation for responsible AI development, governance, and use in the AI RMF, and has collaborated with the public and private sector to implement this voluntary framework. Non-English-speaking nations may not have the resources to develop this type of comprehensive framework. Translating the AI RMF and related documents into non-English languages will advance US leadership and support global responsible AI governance.

The lawmakers wrote, “Countries all over the world are grappling with challenges related to the recent advances in artificial intelligence and automation.”

“Non-English-speaking organizations and countries around the world are struggling to identify, understand, and mitigate risks associated with advances in artificial intelligence, especially due to the fact that standards and guidelines for these systems are woefully underdeveloped. Therefore, we request that NIST work expeditiously to translate the framework and related documents into multiple languages,” the members wrote.

As Chairman of the Senate Commerce Subcommittee on Communications, Media, and Broadband, Senator Luján has strongly advocated for safety guardrails with the use of Artificial Intelligence and for more investments in non-English languages in the digital space. In June, Senator Luján sent a letter to the National Telecommunications and Information Administration (NTIA) Administrator, Alan Davidson, urging the creation of responsible guardrails around Artificial Intelligence (AI) development, governance, and use. Senator Luján introduced the Language-inclusive Support and Transparency for Online Services (LISTOS) Act to improve multilingual large language models, automated decision-making systems, and content moderation practices online to better protect non-English speaking communities.

The full letter can be found here and below:

Dear Director Locascio,

We applaud the National Institute of Standards and Technology for its efforts to establish a best-in-class, voluntary framework for organizations developing and deploying artificial intelligence to mitigate associated risks. This work is a critical step to supporting meaningful risk-based governance of artificial intelligence systems. We write to request that NIST work with stakeholders and other federal agencies to expeditiously translate the Artificial Intelligence Risk Management Framework (RMF) and related guidance into multiple languages to help non-English speaking peoples, organizations, and nations to address the risks they face when developing and deploying this technology. 

Published in January 2021, the RMF was produced in collaboration with stakeholders across public and private sectors in a widely applauded, consensus-driven, open, transparent, and collaborative process. NIST has also released several critical guidance documents to help organizations implement the framework, including a playbook and documents for identifying and mitigating bias in artificial intelligence systems. As evidenced by the scope of participation in the RMF’s creation and its alignment with international standards, the framework is not meant to only support U.S. efforts to improve artificial intelligence governance.

Countries all over the world are grappling with challenges related to the recent advances in artificial intelligence and automation. The European Union is working on the E.U. Artificial Intelligence Act that in part focuses on risk assessments for artificial intelligence systems.  Mexico introduced legislation this year that encourages the creation of an ethical framework for the development of these technologies.  Many other countries across the globe are considering their own rules. For the framework to be usable by the broader international community, it must be translated into multiple languages.

NIST has set this precedent before with its cybersecurity and privacy frameworks. In both cases, NIST has worked with the Department of State’s Digital Connectivity and Cybersecurity Partnership (DCCP) as well as outside organizations to provide dozens of translations on key documents.  In part due to these efforts, NIST’s cybersecurity guidance enjoys widespread adoption and adaptation by the international community, including in Canada, Japan, and Ukraine. 

Non-English-speaking organizations and countries around the world are struggling to identify, understand, and mitigate risks associated with advances in artificial intelligence, especially due to the fact that standards and guidelines for these systems are woefully underdeveloped. Therefore, we request that NIST work expeditiously to translate the framework and related documents into multiple languages. 

We appreciate your dedication to cultivating trust for artificial intelligence systems and look forward to working with you. If you have any questions, please contact Alan McQuinn of the Committee’s Minority staff at (202) 225-6375 and Aditi Gupta in the Office of Senator Luján at (202) 597-2324.

Sincerely,  

###