Senate legislative efforts unveiled last week are targeting the development of artificial intelligence (AI) technologies from markedly different angles – one to prevent against some of the worst-case instances of AI tech deployment, and the other to make sure that Federal AI development efforts leave plenty of room for private sector participation.

Those efforts join a modest stream of AI legislative activity in the Senate following a series of AI insight forums organized by Senate Majority Leader Chuck Schumer, D-N.Y., last year to serve as an AI issues boot camp for legislators interested in those areas.

Framework For Mitigating Extreme AI Risks

One of last week’s notable developments is not yet a piece of formal legislation, but rather the submission of a discussion framework by Sens. Mitt Romney R-Utah., Jack Reed D-R.I., Jerry Moran R-Kan., and Angus King I-Maine., that focuses primarily on mitigating “extreme risks” of AI development.

“Under this potential framework, the most advanced [AI] model developers in the future would be required to safeguard against four extreme risks – the development of biological, chemical, cyber, or nuclear weapons,” the four senators said in a letter to leaders of the Senate’s AI working group. Those leaders include Sens. Schumer, Mike Rounds, R-S.D., Martin Heinrich, D-N.M., and Todd Young, R-Ind.

The senators offering the framework said it would “establish federal oversight of frontier model hardware, development, and deployment to mitigate AI-enabled extreme risks from biological, chemical, cyber, and nuclear threats.”

They defined “frontier” models as “the most advanced AI models that are still yet to be developed.”

“An agency or federal coordinating body would be tasked to oversee the implementation of these proposed requirements, which would apply to only the very largest and most advanced models,” the senators said, adding those “safeguards would be reevaluated on a recurring basis to anticipate evolving threat landscapes and technology.”

Oversight authorities under the framework could be handled by the departments of Commerce and Energy, an interagency group, or an entirely new government agency, the senators suggested.

“My colleagues and I have spent the last several months developing a framework which would create safeguards and provide oversight of frontier AI models aimed at preventing foreign adversaries and bad actors from misusing advanced AI to cause widespread harm,” Sen. Romney said. “It is my hope that our proposal will serve as a starting point for discussion on what actions Congress should take on AI – without hampering American innovation.”

The senators are looking for comments from the Senate AI working group through May 17.

Future of AI Innovation Act

Elsewhere in the Senate, Sens. Young, Maria Cantwell, D-Wash., Marsha Blackburn, R-Tenn., and John Hickenlooper, D-Colo., introduced the Future of AI Innovation Act on April 18 with the stated aim of laying “the foundation to maintain U.S. leadership in the global race to develop artificial intelligence (AI) and other emerging technologies.”

The senators said the bill would:

  • “Formally” establish the existing AI Safety Institute at the National Institute of Standards and Technology (NIST) “to develop voluntary guidelines and standards with the private sector and federal agencies to promote robust long-term innovation in AI,” including “performance benchmarks, evaluations and clear transparency documentation standards for AI systems, while helping companies and consumers better understand and use AI tools”;
  • Create new AI testbed programs between NIST, the Energy Department, the National Science Foundation, and the private sector “to develop security risk tools and testing environments for companies to evaluate their systems for capabilities and limitations”;
  • Create prize competitions for the private sector that focus on “finding AI solutions through the integration of advanced AI technology and emerging technologies like quantum computing”;
  • Accelerate AI innovation by directing Federal science agencies to make available to the public “curated data sets” to accelerate AI advances; and
  • Create a coalition with U.S. allies to cooperate on global AI standards and multilateral research collaborations.

“The Future of AI Innovation Act is critical to maintaining American leadership in the global race to advance AI,” said Sen. Young. “One of my top priorities for federal AI policy is to ensure these technologies are developed in a manner that reflects our democratic values and supports innovation continuing to flourish in the United States, and this bill represents an important step forward in that effort,” the senator said.

Read More About
About
Jose Rascon
Jose Rascon
Jose Rascon is a MeriTalk Staff Reporter covering the intersection of government and technology.
Tags