Make America fake again?: Banning deepfakes of federal candidates in political advertisements under the First Amendment

Objective : to analyze a bill on changes in the US election legislation referring to prohibiting the use of deepfakes of federal candidates in political advertising and to solving certain legal problems based on this research. Methods : the article uses the general dialectical method of cognition, as well as general scientific (analysis, synthesis, induction, deduction) and specific scientific (formal-legal) research methods. Results : The paper considers whether the Protect Elections from Deceptive AI Act, a proposed federal ban of AI-generated deepfakes portraying federal candidates in political advertisements, is constitutional. The author concludes that the bill is constitutional under the First Amendment and that less speech restrictive alternatives fail to address the risks of deepfakes. The paper suggests revisions to narrow the bill's application and ensure its apolitical enforcement: a) the statutory requirement that an individual acts with the purpose of “influencing an election or soliciting funds” should be more specific to hold bad actors accountable; b) the ban should be limited to a fixed period before the election to narrow its application and focus on the most damaging advertisements; c) enforcement by an independent agency, such as the FEC, should be considered to mitigate concerns of partisan enforcement. Scientific novelty : the analysis of the proposed changes in the US election legislation referring to prohibiting the use of deepfakes of federal candidates in political advertising shows that in recent years, artificial intelligence (AI) technology has developed rapidly. Accompanying this advancement in sophistication and accessibility are various societal benefits and risks. For example, political campaigns and political action committees have begun to use AI in advertisements to generate deepfakes of opposing candidates to influence voters. Deepfakes of political candidates interfere with voters' ability to discern falsity from reality and make informed decisions at the ballot box. Deepfakes pose a threat to the integrity of elections and the existence of democracy. Despite the dangers of deepfakes, regulating false political speech raises significant First Amendment questions. Practical significance : the main provisions and conclusions of the article can be used in scientific, pedagogical and law enforcement activities when considering the issues related to regulation of the US elections in particular and the US political system in general.