Login | October 13, 2024
Federal regulators use old laws for new AI tech
RICHARD WEINER
Technology for Lawyers
Published: October 11, 2024
AI fakes are dominating the internet and warping the political, social and economic conversation while governments seem unable to stop them.
While the states, and particularly California, are looking at passing new legislation to try to curtail the negative sides of AI, the federal agencies that have to police AI are announcing that Congress is probably never going to tackle the issue—at least in the near term.
As of now and for the foreseeable, “[w]e have the tools that we have,” according to Consumer Financial Protection Bureau director Rohit Chopra at the Aspen Cyber Summit in D.C. in mid-September.
Federal Communications Commission chair Jessica Rosenworcel said that the FCC has used existing law to deal with AI’s threats. As an example, she discussed the phone call using an AI fake of President Joe Biden to New Hampshire residents.
In that case, the agency applied the 1991 Telephone Consumer Protection Act, which prohibits users generally from using “an artificial or pre-recorded voice to deliver a message.”
But even with that, Rosenworcel recognized that the Biden fake-out was “an early warning shot…. I don’t want to live in a world where we don’t have laws that prevent just this kind of voice cloning and fraud intended to manipulate and election”
Well, Jessice, you may live in that world already, sorry to say. Also I warned all of you about this a decade ago, but whatever.
So, given that federal agencies are left with using their existing, out-of-date laws and regulations to stop AI fakes from impinging on elections or other parts of American lives, agencies should use the principles of disclosure and transparency, Rosenworcel said.
That basically means that any advertising using AI needs to have that fact disclosed in the ad.
The approach also needs the cooperation of the big players in the AI world. FTC chair Lina Khan noted that the innovations necessary to curb the potentially negative effects of AI in advertising come from the private sector.
Of course, the problems are created in the private sector.
Maybe the states, particularly California, can help out, especially since Silicon Valley is a particularly San Francisco-based phenomena. And there are AI bills on Newsone’s desk.
So we shall see.
Because if you think 2024 is bad, wait till you see 2028.
Thanks to Maydeen Merino of the National Law Journal for the analysis.