United States - South Dakota - Deepfake Regulation (HB 1074)

Overview

While the user's request referred to "South Dakota HB 1074 - Deepfakes in Elections," research indicates that the primary legislation addressing deepfakes in South Dakota elections is actually South Dakota Senate Bill 164 (2025), officially titled "An Act to prohibit the use of a deepfake to influence an election and to provide a penalty therefor." This bill was signed into law by Governor Larry Rhoden on March 25, 2025, and is expected to become legally effective on July 1, 2025. The law aims to safeguard the integrity of elections by regulating the use of artificially intelligent (AI) generated or manipulated media, commonly known as deepfakes, in political campaigns. It specifically targets the dissemination of such deceptive content within a critical 90-day window before an election, establishing both civil and criminal liabilities for violations. The enactment of this law reflects a growing national and international concern over the potential for generative AI to mislead voters and undermine democratic processes, marking a significant step for South Dakota in adapting its legal framework to emerging technological challenges.

Definitions

The South Dakota deepfake legislation provides a precise definition for the term "deepfake" to ensure clarity and scope of its application. According to Section 1 of the Act, a "deepfake" is defined as "any image, audio recording, or video recording created or manipulated with the use of artificial intelligence or other digital technology that is so realistic, a reasonable person would believe it depicts the speech or conduct of an actual individual who did not in fact engage in the speech or conduct." This definition is crucial as it establishes two key components: first, the involvement of advanced technological means like artificial intelligence or other digital manipulation, and second, the standard of realism – that a reasonable person would be deceived into believing the content is authentic. The term "materially deceptive" is implicitly addressed by this realism standard, focusing on the potential for the content to mislead the average viewer or listener about an individual's actual actions or statements. The law's focus on "speech or conduct" covers a broad range of potential misrepresentations, from fabricated statements to simulated actions, ensuring comprehensive coverage against various forms of deceptive synthetic media in the electoral context.

Governance and Institutional Framework

The enforcement and oversight of South Dakota's deepfake election law primarily rest within the state's existing judicial and prosecutorial systems. While no new dedicated agency or authority has been established specifically for deepfake regulation, several entities play critical roles. The Attorney General of South Dakota is empowered to seek injunctive or other equitable relief to prohibit the dissemination of deepfakes that violate the Act. This allows for swift legal action to prevent the spread of potentially damaging content before an election. Additionally, candidates who are injured or likely to be injured by a deepfake, as well as individuals depicted in a deepfake, can also initiate civil actions to seek injunctive relief. The responsibility for prosecuting criminal violations (Class 1 misdemeanors) falls to county or state prosecutors, who would bring charges in the state's criminal courts. Civil liability cases, including claims for damages, reasonable costs, and attorney fees, are handled through the state's civil court system. The judiciary, therefore, serves as the ultimate arbiter, interpreting the law's provisions, assessing evidence, and imposing penalties or remedies as appropriate. There are no specific provisions for a regulatory board or commission to oversee compliance, meaning enforcement is primarily reactive, triggered by complaints or investigations rather than proactive monitoring by a specialized body.

Key Focus Areas

• Prohibition on Undisclosed Deepfakes: The law prohibits any person, with the intent to injure a candidate or influence the result of an election, from disseminating a deepfake or entering into a contract to disseminate one, within 90 days of an election. This prohibition applies if the person knows or reasonably should know that the item being disseminated is a deepfake and does not include the required disclosure. This narrow window before an election is considered critical due to the heightened impact of information on voter decisions and the limited time for fact-checking and rebuttal.
• Mandatory Disclosure Requirements: To ensure transparency, the Act mandates specific disclosure language for deepfakes that fall under its scope. The required disclosure states: "This (image/video/audio) has been digitally generated or manipulated and is not an accurate representation of fact or actual events." For images or video recordings, this text must be easily readable by the average viewer, no smaller than the largest font size of any other text in the content, and superimposed over each deepfake. For audio recordings, the disclosure must be clearly spoken in an audible pitch at both the beginning and end of the recording. These requirements aim to provide clear notice to the audience about the manipulated nature of the content.
• Exemptions from Liability: The legislation includes several important exemptions to balance free speech considerations with the need for electoral integrity. These exemptions include deepfakes that constitute satire or parody. Additionally, radio or television broadcasting stations, or cable or satellite television operators, programmers, or producers are exempt if they broadcast a deepfake as part of a bona fide newscast, news interview, news documentary, or on-the-spot coverage of a bona fide news event, provided the broadcast clearly acknowledges, in an easily audible or readable manner, that there are questions about the deepfake's authenticity. They are also exempt if they are paid to broadcast a deepfake and have made a good faith effort to establish that the depiction is not a deepfake. Internet websites or regularly published newspapers, magazines, or other periodicals of general circulation (including internet or digital versions) are also exempt if the deepfake is not paid for by a candidate or political committee, or if it includes the required disclosure.
• Civil Remedies: A person who disseminates a deepfake in violation of the Act is liable to the candidate and the individual depicted in the deepfake for damages, reasonable costs, attorney fees, and any other relief the court deems proper. The plaintiff in such civil actions bears the burden of establishing that the person knowingly made use of a deepfake by clear and convincing evidence, setting a high bar for proving civil liability.
• Injunctive Relief: Beyond civil damages, the Attorney General, an injured candidate, or the individual depicted in a deepfake may seek injunctive or other equitable relief to prohibit the dissemination of the deepfake. This provision allows for proactive measures to stop the spread of deceptive content, which can be critical in the fast-paced environment of an election campaign.

Implementation Framework

South Dakota Senate Bill 164 (2025) is scheduled to become effective on July 1, 2025. As the law does not establish a new administrative body, its implementation will primarily rely on the existing legal and judicial infrastructure of the state. Law enforcement agencies will be responsible for investigating potential criminal violations, while the Attorney General's office will lead efforts for injunctive relief in certain circumstances. Political campaigns, media organizations, and individual citizens are expected to familiarize themselves with the new requirements, particularly regarding the 90-day pre-election window and the specific disclosure mandates. The law's integration into the South Dakota Codified Laws, likely within Chapter 12-26 concerning election offenses, means it will be enforced similarly to other election-related statutes. No specific timelines for phased implementation or educational campaigns by state agencies are outlined in the bill, placing the onus of compliance largely on the entities and individuals involved in political communication. Given the novelty of deepfake regulation, it is anticipated that early cases may help establish precedents and clarify interpretations of the law's provisions, particularly around the "reasonable person" standard for deception and the "good faith effort" for media exemptions.

Monitoring and Evaluation

Monitoring and evaluation of compliance with South Dakota's deepfake election law will primarily be conducted through a combination of public vigilance, complaints from affected parties, and the investigative actions of law enforcement and the Attorney General's office. The law does not establish a dedicated state agency or mechanism for proactive monitoring of digital media for deepfake content. Instead, it relies on candidates, political committees, and the general public to identify and report potential violations. The provisions allowing injured candidates and depicted individuals to seek civil remedies and injunctive relief serve as a primary enforcement mechanism, incentivizing private action to ensure compliance. The Attorney General’s authority to seek injunctive relief also provides a state-level avenue for intervention. Courts will play a central role in evaluating claims, interpreting the definition of a deepfake, assessing intent, and determining whether appropriate disclosures were made. While no formal reporting requirements are imposed on campaigns or media outlets, the threat of criminal prosecution and civil liability acts as a deterrent. The effectiveness of the law will likely be evaluated over time through the number of complaints filed, cases prosecuted, and civil judgments rendered, as well as through public awareness and the observed prevalence of undisclosed deepfakes in future South Dakota elections. Amendments or further legislative action may be considered based on practical experience with the law's enforcement and its impact on electoral discourse.

Penalties, Liability, and Appeals

South Dakota's deepfake election law imposes significant penalties and establishes clear liability for violations. Any person who disseminates a deepfake in violation of the Act – specifically, with the intent to injure a candidate or influence an election within 90 days of an election, knowing or reasonably knowing it's a deepfake without the required disclosure – commits a Class 1 misdemeanor. In South Dakota, a Class 1 misdemeanor carries a maximum penalty of one year in county jail and/or a fine of $2,000. This criminal penalty underscores the seriousness with which the state views the intentional misuse of deepfake technology in political contexts. Beyond criminal charges, the law also establishes civil liability. A person who violates the Act is liable to the candidate and the individual depicted in the deepfake for actual damages incurred, reasonable costs, and attorney fees. Courts also have the discretion to grant any other relief deemed proper. The burden of proof for establishing civil liability rests with the plaintiff, who must demonstrate by clear and convincing evidence that the defendant knowingly made use of a deepfake in violation of the statute. This higher standard of evidence in civil cases (beyond a mere preponderance of evidence) reflects the potential impact on free speech rights. Appeals from both criminal convictions and civil judgments would follow the standard appellate procedures within the South Dakota court system, typically moving from circuit court to the South Dakota Supreme Court.

Relationship to Other Instruments

South Dakota Senate Bill 164 (2025) operates within the existing framework of state and federal law, particularly concerning elections, campaign finance, and free speech. It supplements South Dakota's existing election laws, adding a new category of prohibited election interference related to deceptive media. The law's provisions must be interpreted in light of the First Amendment to the United States Constitution, which protects freedom of speech. The inclusion of exemptions for satire, parody, and bona fide news reporting, along with the requirement for intent to injure or influence an election and the 90-day window, are likely designed to withstand constitutional scrutiny by balancing the state's compelling interest in fair elections against free speech rights. The "reasonable person" standard and the "knowing or reasonably should know" standard for intent are crucial in this regard, aiming to target genuinely deceptive content rather than accidental misrepresentations or protected speech. While the law primarily addresses state elections, its principles could interact with federal campaign regulations if deepfakes are used in federal election campaigns within South Dakota. It also aligns with a broader trend among U.S. states to address AI-generated content in elections, indicating a recognition of the novel challenges posed by this technology that existing laws, such as those against defamation or false advertising, may not fully cover due to the unique nature of synthetic media.

International Alignment

South Dakota's deepfake election law aligns with a growing international and national movement to regulate the use of artificial intelligence in political discourse, particularly concerning its potential to create deceptive content. While there isn't a specific international treaty or universal standard for deepfakes in elections, many democratic nations and sub-national jurisdictions are grappling with similar legislative efforts. Within the United States, South Dakota joins a significant number of states that have enacted or are considering legislation to address election deepfakes, including Texas, California, Washington, and Michigan, among others. These state-level initiatives often share common elements, such as defining deepfakes, requiring disclosures, establishing prohibitions within certain pre-election periods, and outlining penalties. Globally, discussions are ongoing within bodies like the European Union (e.g., the AI Act, which includes provisions for transparency of AI-generated content) and various national governments about how to manage the risks posed by generative AI. South Dakota's approach, focusing on transparency through disclosure and penalizing malicious intent, reflects a common regulatory philosophy seen in other jurisdictions: to enable the use of AI tools while mitigating their potential for harm, particularly in sensitive areas like democratic elections. The law contributes to the evolving landscape of AI governance, positioning South Dakota as part of a broader effort to establish guardrails for this rapidly advancing technology.

Implementation Timeline

Compliance Checklist

Sources and References