Representative Laurel Lee (R-FL) discussed the need for a federal standard on the regulation of artificial intelligence (AI) in a recent House Judiciary hearing.
Rep. Lee asked hearing witness Kevin Frazier, AI Innovation and Law Fellow at the University of Texas School of Law, to elaborate on the reference he made to Louis Brandeis's "laboratories of democracy" concept in relation to states' attempts to impose regulations on AI.
Frazier replied that Brandeis's "laboratories of democracy" phrase has the forgotten portion of, "you can run an experiment without risk to the nation [author's emphasis]." California's AI regulations, he posited, "do pose a risk to the nation because they try to impede AI innovation itself."
Thus, Frazier argued that AI companies do not want to change their models repeatedly to comply with various state laws, thereby halting innovation altogether.
Rep. Lee questioned how the current copyright law "operates in really ensuring that we are still honoring those concepts of content creators and intellectual property," inquiring how the AI and copyright laws can be harmonized.
Frazier noted that in copyright disputes, settlements primarily go to the publishers, not the author. Thus, "if we are trying to incentivize the creation of new art, new discoveries, and new scientific discoveries, copyright may not be the vehicle needed."
He further suggested that because current copyright law is 70 years plus the life of the author, "the fact that we have ended up in a world in which a handful of publishers may be able to dictate the quality of our AI, I think, is antithetical to the original purpose of the IP clause."
The Florida Congresswoman also questioned R Street Institute's Senior Technology & Innovation Fellow Adam Thierer.
"One of the things that you touched on was the idea that Congress could explore giving [the National Institute of Standards and Technology (NIST)] or [the Center for AI Standards and Innovation (CAISI)] more authority to develop standards for AI frontier models? Should we designate a single federal entity to try and develop those standards?" she questioned.
Thierer responded that the NIST or CAISI's authority is necessary because "many states are attempting to impose a very technocratic type of design on artificial intelligence models and systems preemptively in an almost European-style way, and it is a huge problem."
Not only did Thierer say that patchwork laws stifle innovation for reasons similar to those described by Frazier, but "these states lack the technical capability to do some of this in certain circumstances and lack the information needed to do it properly."
