WASHINGTON — Utah Sen. Mitt Romney and several Senate colleagues announced Tuesday a blueprint to protect against potential threats posed by artificial intelligence — including biological, chemical, cyber and nuclear threats — and future advances in technology. By increasing federal oversight of .
A bipartisan group of senators — including Romney, a Republican, Sens. Jack Reed, D-Rhode Island, Jerry Moran, R-Kansas, and Angus King, I-Maany — outlined their framework in a letter to Senate leaders. Outlined. The Artificial Intelligence Working Group’s plan addresses so-called frontier models that have yet to be developed, and would only apply to models designed with a wide range of capabilities or “bioengineering.” , intended to be used for chemical engineering, cyber security or nuclear development”, according to a fact sheet. .
AI has the potential to dramatically improve and change the way we live, but it comes with enormous risks to national security and our humanity at large,” Romney said. My colleagues and I have spent the past several months developing a framework that will create safeguards and monitor frontier AI models that are designed to advance foreign adversaries and bad actors to cause widespread harm. AI is to be prevented from being misused.
It’s my hope that our proposal will serve as a starting point for a discussion about what steps Congress should take on AI — without stifling innovation.
Although artificial intelligence only emerged into the public consciousness less than two years ago with the release of ChatGPT, policymakers around the world are already grappling with what, if any, impact the rapidly improving technology has. What precautions should be taken? The European Union approved the first major set of regulations last month, which sees the use of artificial intelligence as a higher risk under increased scrutiny in the 27-nation bloc.
US leaders have yet to adopt comprehensive AI regulations, but Congress has held several panels and sessions with tech leaders, such as OpenAI CEO Sam Altman, to address the technology. During one such Senate hearing last September, Romney said he was “in the camp of being more fearful of AI than in the camp of people who think it’s going to make the world better.”
President Joe Biden signed an executive order in October that focuses on AI security and privacy for Americans.
AI regulation is also a concern at the state level, and Utah lawmakers created an artificial intelligence learning lab earlier this year to help state government keep abreast of the latest opportunities and risks posed by advancing technology. Get help.
The senators’ letter for AI regulation contains few specifics, rather than pursuing a broad regulatory framework. Similar to European regulations, the framework proposes a tiered licensing structure where high-risk artificial intelligence models can be “licensed for designated users or limited public use” rather than being deployed generally.
The proposal suggests several options for overseeing AI, including the Commerce Department’s National Institute for Standards and Technology and the Bureau of Industry and Security, or US National Laboratories located in the Department of Energy. Congress may also choose to create a new federal agency or a new interagency coordinating body to work among existing agencies.
Regardless of who oversees the regulations, the senators suggest that any oversight body include “subject matter experts who may be detailed from relevant federal agencies” and “talented AI scientists and engineers.
Romney and his colleagues acknowledge in the letter that their framework is only intended to “serve as a basis for international coordination to mitigate the extreme threats posed by AI,” and that congressional leaders And other stakeholders “encourage further discussions in the coming weeks. That will be necessary for us to create a framework that can become law.”