Can Existing Election Laws Address AI-Generated Misinformation?

With the rise of generative Artificial Intelligence (AI), people have the ability to generate any information at the touch of a button. AI provides a lot of shortcuts—research that previously took hours can now take minutes. This applies to all aspects of research, including candidate research prior to an election. However, with this increased ability, people also become more vulnerable to misinformation. The increased reliance on AI for research causes people to cut corners when it comes to fact checking claims. When it comes to election research, AI leaves people vulnerable to election misinformation, which can lead people to cast a vote based on misinformation. While AI is prone to mistakes, many people treat its output as truth since they view it as an omniscient source of information. Because of people's vulnerability to AI misinformation, election laws should address the rise in AI shortcuts people take and make sure that users are clearly informed about the possibility of false information. This paper is being divided into three sections. Section I provides a background and existing legal landscape for potential protections on AI generated content. Section II will be an analytical critique of the existing laws. It will analyze how the laws apply to AI generated content, and whether those laws are feasible for the scope of AI content. It will also cover theoretical implications of the growth of AI content under the existing legal framework. Finally, Section III will propose some solutions, under the assumption that AI content will continue to improve.

Background / Existing Legal Landscape

Historical Development of Election Speech Doctrine

The rise of generative AI raises new questions about whether existing election laws can effectively address misinformation while remaining consistent with the First Amendment. Political speech has long received the highest level of constitutional protection because it is essential to democratic self-government. The Supreme Court has recognized that speech regarding politics, public affairs, and elections lies at the core of the First Amendment and generally cannot be restricted merely because it is controversial or offensive.

These protections extend beyond traditional political advocacy. In McIntyre v. Ohio Elections Commission, the Court held that anonymous political speech is protected by the First Amendment, emphasizing the historical importance of anonymous participation in public debate. The Court has also been reluctant to regulate false political speech. In New York Times Co. v. Sullivan, the Court acknowledged that false statements are sometimes "inevitable in free debate." In United States v. Alvarez, the Court held that false statements do not automatically lose constitutional protectionn. These cases are an example of an extensive legal precedent in regards to political speech protections. Together, they create a significant constitutional barrier in addressing AI generated election misinformation.

Current Statutory Framework

Although election speech receives broad constitutional protection, federal and state governments have enacted laws aimed at preventing voter deception. At the federal level, the Federal Election Campaign Act (FECA) prohibits fraudulent misrepresentation by candidates and campaigns and requires political advertisements to disclose who paid for the communication. The Federal Election Commission has clarified that these provisions are technology-neutral and may apply to deceptive AI-generated content, although enforcement currently occurs on a case-by-case basis.

States on the other hand, have taken a more aggressive approach. As of 2026, 30 states have enacted laws regulating political deepfakes. Most states require disclosures that inform viewers when political advertisements contain AI generated or substantially altered content. However, a smaller number of states prohibit certain deepfakes within a specified time period prior to an election. Despite these efforts, constitutional challenges have been a limiting factor in getting these laws to remain intact. Courts have criticized laws that broadly restrict political deepfakes, finding that vague standards may suppress protected political speech.

Emerging AI Election Challenges

AI generated misinformation presents challenges that go beyond traditional forms of political deception. Deepfakes, synthetic audio, and AI image manipulation can create fabricated content that appears realistic to the unsuspecting eye. For example, during the 2024 election cycle, AI generated images used in campaign messaging and robocalls to voters that imitated former President Biden's voice were used to discourage voter participation in the New Hampshire primary.

The speed with which online information travels further amplifies these risks. False information, especially information that contains controversy, often spreads faster than fact checkers can respond, allowing misleading content to influence voters before corrections can be issued. Unlike traditional misinformation, AI generated content can be produced rapidly, distributed widely, and made increasingly difficult to distinguish from authentic media. These developments raise questions about whether existing election laws are equipped to address the unique challenges posed by generative AI.

Analytical Critique

Why the current framework fails

Although existing election laws provide provisions for addressing voter deception and election fraud, they were designed for traditional forms of misinformation rather than AI-generated content. As a result, these laws often struggle to address the speed, scale, and anonymity that characterize modern AI-generated misinformation.

One challenge is identifying the source of deceptive content. AI tools allow users to create realistic images, videos, audio recordings, and text with little technical expertise; while the ability to create anonymous accounts, delete histories, and enable temporary chats can make creators difficult to trace. This problem is exacerbated by constitutional protections for anonymous political speech, which limit the government's ability to regulate online communications without raising First Amendment concerns. Jurisdictional issues create additional barriers, as AI-generated content intended to influence voters in one state may be created in another state or even another country.

The rapid spread of AI-generated content further limits the effectiveness of existing laws. Misleading content can be distributed across multiple social media platforms within minutes, reaching voters before election officials or fact-checkers have an opportunity to respond. Although legal remedies may eventually result in penalties or the removal of deceptive content, elections occur on short timelines while litigation often takes months or years to resolve. Consequently, misinformation may influence voter decisions long before legal action becomes an option.

Finally, advances in generative AI have made deceptive content increasingly difficult for voters to identify. Research has shown that exposure to AI-generated media can increase uncertainty regarding the authenticity of information, while the low cost of AI tools allow large volumes of content to be produced and distributed rapidly. Together, these characteristics suggest that laws designed to address human-generated misinformation lack the capacity to address the unique challenges posed by AI-generated election content.

Constitutional Conflict

First Amendment protections for political speech complicate efforts to regulate AI generated election misinformation. As previously mentioned, political speech receives the highest level of constitutional protection, even when it contains false or misleading information. In Citizens United v. FEC, the Supreme Court held that political spending by corporations and labor unions constitutes protected political speech, demonstrating the broad scope of First Amendment protections in the electoral context. As a result, attempts to regulate AI-generated political content may be viewed as restrictions on protected speech rather than efforts to prevent misinformation.

The Constitution also places significant limits on content-based regulations. In Reed v. Town of Gilbert, the Supreme Court held that content-based restrictions are subject to strict scrutiny. Consequently, laws targeting AI-generated political content must be narrowly tailored to serve a compelling government interest, creating substantial barriers to broad restrictions on election-related deepfakes.

Supporters of the current framework argue that existing fraud and election-interference laws are sufficient to address deceptive AI content. However, these laws were developed before the emergence of generative AI and often require proof of intent, causation, or identifiable harm, which may be difficult to establish when content is created anonymously and distributed rapidly online. Others contend that misinformation can be corrected through counterspeech and fact-checking. While this is still important to free political discourse, AI-generated content can spread to large audiences within minutes, often reaching voters before corrections can be issued. These challenges suggest that existing legal protections for political speech, while essential to democratic debate, may also limit the effectiveness of current efforts to address AI-generated election misinformation.

Proposed Solution

Although existing election laws provide some protection against voter deception, they were not designed to address the unique challenges posed by AI-generated content. Rather than imposing broad restrictions on political speech, lawmakers should adopt disclosure based regulations similar to those already used in campaign finance law. Under the FECA, political advertisements must disclose who paid for the communication. A comparable framework could require political content that is created or substantially altered by artificial intelligence to include a clear disclaimer informing viewers that AI was used. Such disclosures would allow voters to make their own judgments regarding the credibility of the information while avoiding many of the constitutional concerns associated with outright bans on political speech.

In addition to disclosure requirements, lawmakers could require AI companies to embed watermarks or metadata identifying AI-generated content. Because these measures focus on transparency rather than censorship, they are more likely to withstand First Amendment scrutiny while helping voters distinguish authentic media from manipulated content.

However, these solutions are not without limitations. Enforcement would remain difficult because deceptive content can spread rapidly across multiple platforms before regulators have an opportunity to intervene. Users may also remove watermarks or repost AI-generated content without required disclosures. Furthermore, litigation and regulatory investigations often move much more slowly than election cycles, meaning misinformation may influence voters before legal remedies become available. While disclosure requirements cannot eliminate AI-generated election misinformation, they offer a practical and constitutionally viable approach for reducing its impact while preserving protections for political speech.

Conclusion

The rise of generative artificial intelligence has created significant gaps in the existing legal framework governing election misinformation. Although current election laws provide protections against fraud and voter deception, they were developed for traditional forms of communication and struggle to address the speed, anonymity, and scale of AI-generated content. At the same time, the First Amendment places substantial limits on government efforts to regulate political speech, making broad restrictions on AI-generated election content constitutionally difficult to sustain.

The question is not whether AI will continue to influence elections, but whether legal institutions can adapt quickly enough.

As AI technology continues to improve, the challenge for lawmakers will be balancing election integrity with the constitutional commitment to free and open political debate. While no legal framework can completely eliminate misinformation, disclosure-based regulations provide a practical middle ground. By requiring transparency regarding the use of artificial intelligence in political communications, lawmakers can help voters evaluate the credibility of election-related information without imposing broad restrictions on protected speech. Ultimately, the question is not whether AI will continue to influence elections, but whether legal institutions can adapt quickly enough to ensure that voters remain informed in an era where digital content is increasingly difficult to distinguish from reality.