In a motion raised during a recent Los Angeles City Council meeting, 10th District Councilmember Heather Hutt has called for stricter regulations on the use of artificial intelligence, deep fakes and other technological abuses in political campaigns for city office positions. She has also asked for the city council to request that the City Ethics Commission, in consultation with the Office of the City Attorney, report back with recommendations on how to ensure that all campaign communications available on the commission’s website are encrypted with a watermark so that documents posted cannot be abused by opponents to create false claims of ethics violations, generate deep fakes or cause confusion for voters and the media through any other means.
“In the burgeoning era of rapid and wide use of artificial intelligence and evolving technology, it is inevitable that the various image, video and other documents available on the internet will be abused in a variety of ways,” Hutt wrote in the motion, which was seconded by LA City Council President Paul Krekorian and District 11 Councilmember Traci Park. “One of the ways we will see more and more of in 2024 and beyond is the perverse use of this information by political opponents, independent expenditures and other nefarious individuals and entities to confuse voters, the media and regulatory agencies like the City Ethics Commission, wasting valuable time and resources on false information and false ethics violation claims.”
According to Hutt, the intellectual property of political campaigns and campaign communications currently available on the City Ethics Commission’s website are unprotected and available to download and edit by any individual who has the software capability to do so.
“In order to protect the integrity of political campaigns in the City of Los Angeles, the City Ethics Commission must develop recommendations that will regulate these technological advances in a manner that gives voters in the City of Los Angeles access to the best and most accurate information to inform their decisions in the voting booth,” Hutt said in her motion.
AI-generated deepfakes have been used to spread false information ahead of elections in the past, most recently in New Hampshire.
On Jan. 21, two days before the New Hampshire Presidential Primary Election, a robocall using AI-generated audio to mimic President Joe Biden’s voice and style of language was sent to thousands of New Hampshire voters.
According to the New Hampshire Department of Justice, recipients of the call were encouraged not to vote in the Jan. 23 presidential primary election, saying “your vote makes a difference in November, not this Tuesday.”
The software company ElevenLabs’ voice-cloning technology was reportedly used to generate the audio and has since banned the account involved. The Federal Communications Commission has now made it illegal to produce robocalls using AI-generated voices.
“AI-generated recordings used to deceive voters have the potential to have devastating effects on the democratic election process,” Attorney General John M. Formella said.
