WASHINGTON— Senator Mitt Romney from Utah, alongside fellow Senate colleagues, has introduced a proposal aimed at enhancing national supervision of emerging cutting-edge technologies to safeguard against potential threats posed by various forms of artificial intelligence, including biological, chemical, computer, and nuclear risks.
The bipartisan group of senators, which includes Romney, a Republican, as well as Senators Jack Reed (D-Rhode Island), Jerry Moran (R-Kansas), and Angus King (I-Maine), presented their initiative to bolster oversight of frontier technologies to counter potential risks. The initiative focuses on models that are yet to be developed and would specifically target models with diverse capabilities or those designed for bioengineering, chemical engineering, security, or nuclear enhancements, as outlined in a statement.
Romney emphasized the dual nature of artificial intelligence, stating that while AI holds the potential to revolutionize various aspects of society, it also poses significant risks to national security and humanity as a whole. He highlighted the importance of establishing safeguards and regulatory mechanisms to prevent adversaries and malicious actors from exploiting advanced AI technologies and causing widespread harm.
The proposed framework aims to stimulate discussions within Congress on the appropriate approach to regulating AI without stifling innovation. It comes at a time when policymakers globally are grappling with the need for regulatory frameworks to govern the rapid advancements in technology. Despite the recent mainstream recognition of technologies like ChatGPT, comprehensive AI regulations are yet to be adopted in the United States, although there have been congressional engagements with industry leaders to address AI-related concerns.
President Joe Biden’s executive order focusing on American AI safety and privacy underscores the growing recognition of the importance of AI governance at the federal level. In Utah, the establishment of the Artificial Intelligence Learning Laboratory reflects a proactive approach to staying informed about the opportunities and risks associated with technological advancements.
The senators’ proposal outlines a broad regulatory framework, suggesting the implementation of a tiered licensing system for high-risk AI models, potentially restricting their use to vetted customers or limited public deployment. The proposal also recommends the involvement of key federal agencies such as the National Institute for Standards and Technology, Bureau of Industry and Security, and Department of Commerce’s National Institute for Standards and Technology, as well as considering the establishment of a new federal agency or interagency coordination body to oversee AI regulations effectively.
Romney and his colleagues acknowledge that their framework serves as a foundational basis for international collaboration to mitigate the severe risks posed by AI and call for further discussions among congressional leaders and stakeholders to refine the framework into actionable legislation.
