United States - South Dakota - Deepfakes Regulation (SB 164)

South Dakota SB 164 - Deepfakes Regulation

United States

RAI-US-SD-SDS1DXX-2025
Effective: March 31, 2025
In Force(In Force)
ActTransparency and DisclosureEnforcement and PenaltiesFundamental Rights
Export PDF

South Dakota SB 164 prohibits undisclosed deepfakes in elections, imposing penalties for malicious use and requiring transparency to protect electoral integrity.

Overview

South Dakota Senate Bill 164, now enacted into law, establishes a regulatory framework specifically targeting the use of deepfakes in state elections. The primary objective of this legislation is to prevent the spread of deceptive and manipulated media that could influence voters and undermine the integrity of the electoral process. Enacted on March 31, 2025, the Act prohibits any person from disseminating a deepfake with the intent to injure a candidate within 90 days of an election, unless the deepfake includes a specific, prominent disclosure. This measure addresses growing concerns over the potential for artificial intelligence and other digital technologies to create highly realistic but fabricated content, such as images, audio recordings, or video recordings, that falsely depict individuals, particularly political candidates, engaging in speech or conduct that never occurred.

The scope of South Dakota SB 164 is focused narrowly on election-related deepfakes, distinguishing it from broader deepfake regulations that might cover other areas like non-consensual intimate imagery. By setting a 90-day pre-election window, the Act aims to mitigate last-minute disinformation campaigns that could significantly impact election outcomes without sufficient time for factual correction. It introduces both civil and criminal penalties for violations, underscoring the state's commitment to deterring malicious use of deepfake technology. Furthermore, the Act provides clear guidelines for what constitutes an acceptable disclosure, offering an affirmative defense for those who use deepfakes for legitimate purposes, such as satire, or with proper labeling. This balanced approach seeks to protect free speech while simultaneously safeguarding democratic processes from technological manipulation.

Definitions

South Dakota SB 164 introduces crucial definitions into state law to precisely delineate the scope of its deepfake regulations. For the purposes of this 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. This manipulated content must either appear to a reasonable person to depict a real individual saying or doing something that did not actually occur in reality, or provide a reasonable person with a fundamentally different understanding or impression of the actual events. This definition is critical as it establishes a standard of realism and potential for deception that the manipulated media must meet to fall under the Act's purview, ensuring that only genuinely misleading content is targeted.

The Act also provides a definition for "artificial intelligence," characterizing it as machine-based systems capable of generating content or outputs. This broad definition acknowledges the evolving nature of AI technology and ensures that the legislation remains relevant as AI capabilities advance. By explicitly defining both "deepfake" and "artificial intelligence," South Dakota SB 164 provides a clear legal foundation for enforcement and compliance, helping to avoid ambiguity in its application. These definitions are central to understanding the Act's intent to regulate technologically advanced forms of media manipulation that pose a threat to election integrity, while also distinguishing them from traditional forms of media editing or satire.

Governance and Institutional Framework

South Dakota SB 164 does not establish a new dedicated governmental agency or institutional framework specifically for the oversight of deepfakes. Instead, the enforcement and governance of this Act are integrated into existing legal and judicial structures within the state. The responsibility for addressing violations primarily falls upon the state's prosecutorial authorities, such as the Attorney General or local district attorneys, for criminal offenses. For civil remedies, the judicial system is empowered to hear cases brought by injured candidates or individuals depicted in deepfakes, allowing them to seek damages and other relief. This approach leverages established legal mechanisms, ensuring that the regulation of deepfakes benefits from existing investigative and adjudicative processes.

The Act outlines the roles of various parties in its implementation. While no new regulatory body is created, the legislative intent suggests that the South Dakota Legislature will continue to monitor the impact and effectiveness of this law, potentially leading to future amendments or supplementary legislation as AI technology evolves. The judiciary plays a crucial role in interpreting the Act's provisions, particularly concerning the "reasonable person" standard for deepfake identification and the "clear and convincing evidence" burden of proof for civil liability. This distributed governance model relies on the existing legal infrastructure to uphold the Act's provisions, rather than creating new bureaucratic layers, aiming for efficient and adaptable enforcement within the state's established governmental framework.

Key Focus Areas

The primary focus area of South Dakota SB 164 is the preservation of election integrity by regulating the use of deepfakes in political campaigns. The Act specifically targets the dissemination of manipulated media that is intended to injure a candidate within a critical 90-day period leading up to an election. This timeframe is chosen to address the heightened vulnerability of the electoral process to last-minute disinformation, where false narratives can spread rapidly and significantly influence voter perception without adequate time for rebuttal or correction. The legislation aims to prevent scenarios where AI-generated content could falsely portray candidates in a negative light, thereby distorting public discourse and potentially swaying election outcomes.

Another key focus is on transparency and disclosure. Rather than imposing an outright ban on deepfakes, the Act emphasizes the requirement for clear and conspicuous labeling when such content is used in an election context. This approach acknowledges the potential for legitimate uses of AI-generated media, such as satire or parody, which are explicitly exempted from liability. The disclosure mandate ensures that voters are informed when they are viewing or listening to manipulated content, allowing them to critically assess the information. Furthermore, the Act addresses the protection of individual rights by providing a legal avenue for candidates and depicted individuals to seek redress for harm caused by malicious deepfakes, reinforcing the principle that individuals should not be falsely represented in a manner that damages their reputation or standing.

Implementation Framework

The implementation framework for South Dakota SB 164 centers on a disclosure-based regulatory model, coupled with specific prohibitions and penalties. For an image or video recording deemed a deepfake under the Act, the required disclosure must state: "This (image/video/audio) has been digitally generated or manipulated and is not an accurate representation of fact or actual events." This text must appear in a size that is easily readable by the average viewer and no smaller than the largest font size of other text appearing in the image or video recording. Crucially, the disclosure must be superimposed over each deepfake, ensuring its visibility. These stringent formatting requirements aim to make the disclosure unavoidable and easily comprehensible to the audience, preventing subtle or hidden disclaimers.

For audio recordings, the disclosure requirements are tailored to the medium: the statement must be read in a clearly spoken manner and in a pitch that is easily heard by the average listener. This audible disclosure must be presented at both the beginning and the end of the audio recording. The Act also establishes an affirmative defense, meaning that if a deepfake includes the prescribed disclosure, the disseminator may not be held liable under the Act. This mechanism encourages compliance through transparency. Additionally, the Act provides exemptions from liability for broadcasters, newspapers, websites, and radio stations, provided they are not directly involved in the creation or intentional dissemination of the harmful deepfake, and for content that clearly constitutes satire or parody, balancing regulation with free speech considerations.

Monitoring and Evaluation

As South Dakota SB 164 is a recently enacted piece of legislation, specific formal mechanisms for its ongoing monitoring and evaluation are not explicitly detailed within the Act itself. However, the effectiveness of such a law will inherently be subject to continuous observation through several indirect channels. The primary method of evaluating its impact will likely involve tracking instances of alleged deepfake misuse in elections, the number of civil lawsuits filed, and criminal prosecutions initiated under its provisions. Legal professionals, election officials, and advocacy groups will naturally monitor these cases to assess how the law is being applied by the courts, the clarity of its definitions, and the efficacy of its penalties in deterring harmful deepfake dissemination.

Furthermore, the legislative body, particularly committees focused on government affairs or judiciary, may conduct periodic reviews or hold hearings to gather feedback on the law's operation. This could involve soliciting input from law enforcement, election administrators, legal experts, and the public regarding any challenges or successes encountered during its implementation. The evolving nature of artificial intelligence and deepfake technology also necessitates ongoing legislative attention, implying that the Act may be subject to future amendments or updates based on technological advancements and practical experiences. While no dedicated monitoring body is established, the dynamic interplay of legal enforcement, judicial interpretation, and potential legislative review will serve as the de facto system for monitoring and evaluating the Act's real-world impact on election integrity in South Dakota.

Penalties, Liability, and Appeals

South Dakota SB 164 imposes significant penalties for violations related to the unauthorized dissemination of deepfakes in elections. Any person who, with the intent to injure a candidate, disseminates a deepfake or enters into an agreement to disseminate a deepfake within 90 days of an election, knowing or reasonably knowing it is a deepfake and without the required disclosure, is guilty of a Class 1 misdemeanor. A Class 1 misdemeanor in South Dakota carries potential penalties of up to one year imprisonment in a county jail and/or a fine of up to $2,000. These criminal provisions underscore the seriousness with which the state views the malicious use of deepfake technology to influence electoral outcomes.

In addition to criminal penalties, the Act also establishes civil liability. 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, and attorney fees, as well as any other relief the court deems proper. The plaintiff in such a civil action bears the burden of establishing that the person knowingly made use of a deepfake by clear and convincing evidence. This high evidentiary standard ensures that civil claims are based on substantial proof of malicious intent and knowledge. The Act explicitly states that these civil remedies do not limit any other cause of action that may be brought against a person who disseminates a deepfake, allowing for other legal avenues like defamation claims to be pursued concurrently. The appeals process for both criminal and civil judgments would follow the standard procedures within the South Dakota state court system.

Relationship to Other Instruments

South Dakota SB 164 operates within the existing legal landscape of the state and the broader United States, complementing rather than supplanting other relevant laws. Its primary relationship is with South Dakota's election laws, as it specifically targets deepfakes used to influence elections. The Act adds a new layer of protection to the integrity of the electoral process, working in conjunction with established statutes governing campaign conduct, voter deception, and fair electoral practices. It provides a specific tool to address a modern form of misinformation that traditional election laws might not explicitly cover, thereby strengthening the overall framework for secure and transparent elections.

Furthermore, the Act interacts with general civil and criminal statutes. The civil liability provisions for damages, costs, and attorney fees can be seen as an extension of tort law, providing specific redress for harm caused by deepfakes, potentially alongside existing defamation or invasion of privacy claims. The criminal penalties for a Class 1 misdemeanor align with the state's criminal code. Importantly, the Act includes exemptions for satire, parody, and certain media organizations (broadcasters, newspapers, websites, and radio stations), which reflects an effort to balance the regulation of deepfakes with First Amendment free speech protections and the role of the press. This demonstrates a careful consideration of how the new deepfake regulations fit within the broader constitutional and statutory framework, aiming to avoid conflicts while addressing a specific societal concern.

International Alignment

As a state-level legislative act, South Dakota SB 164 does not directly align with international treaties, conventions, or supra-national regulations. Its jurisdiction is confined to the State of South Dakota, United States, and it is primarily concerned with domestic election integrity. However, the principles underlying the Act—namely, the regulation of AI-generated deceptive content, the emphasis on transparency through disclosure, and the imposition of penalties for malicious use—resonate with broader global discussions and emerging regulatory trends concerning artificial intelligence and deepfakes. Many nations and international bodies are grappling with similar challenges posed by synthetic media, particularly in the context of misinformation, electoral interference, and the protection of individual rights.

While not a direct participant in international alignment efforts, South Dakota's enactment of this law contributes to a growing body of sub-national and national legislation worldwide addressing deepfakes. This collective legislative activity, even at a state level, informs and influences the global discourse on AI governance. Other states in the U.S. and countries globally are exploring similar disclosure requirements, prohibitions on election-related deepfakes, and civil/criminal liabilities. Therefore, while the Act itself is a domestic instrument, its existence reflects and contributes to a worldwide trend of developing legal frameworks to manage the societal impacts of advanced AI technologies.

Implementation Timeline

MilestoneDateNotes
Bill Introduced2025-01-30Senate Bill 164 was formally introduced in the South Dakota Senate.
Committee Substitute Issued2025-02-24A committee substitute for the bill was issued, reflecting amendments or revisions made during the committee review process.
Senate State Affairs Committee Do Pass Amended2025-02-24The Senate State Affairs Committee voted to pass the bill as amended (YEAS 8, NAYS 1).
Senate Second Reading/Final Passage2025-03-01The bill was presented for its second reading and final passage in the South Dakota Senate.
Enrolled Bill Text Available2025-03-11The enrolled version of the bill, reflecting all passed amendments, was made available.
Signed by the Governor2025-03-31Governor Larry Rhoden signed the bill into law (S.J. 539). Note: Some sources indicate March 25, 2025, as the signing date.
Act In Force2025-03-31The Act became effective upon the Governor's signature.

Compliance Checklist

CheckRequired Action
Identify Deepfake ContentDetermine if any image, audio, or video recording created or manipulated with AI or digital technology meets the definition of a "deepfake" under the Act (i.e., appears to depict a real individual saying/doing something that didn't occur or provides a fundamentally different understanding of events).
Assess Intent to InjureEvaluate if the deepfake is disseminated with the intent to injure a candidate. The Act specifically targets malicious intent.
Verify Election ProximityConfirm if the dissemination occurs within 90 days of an election. The Act's prohibitions are time-sensitive to the electoral cycle.
Apply Disclosure for Images/VideosIf disseminating a deepfake image or video, ensure it includes the disclosure: "This (image/video/audio) has been digitally generated or manipulated and is not an accurate representation of fact or actual events." The text must be easily readable, no smaller than the largest font size of other text, and superimposed over the deepfake.
Apply Disclosure for AudioIf disseminating a deepfake audio recording, ensure the disclosure is clearly spoken in an audible pitch at both the beginning and end of the recording.
Consider ExemptionsDetermine if the deepfake falls under any exemptions, such as satire, parody, or dissemination by certain media entities (broadcasters, newspapers, websites, radio stations) not involved in creation or intentional harmful dissemination.
Understand PenaltiesBe aware that non-compliance can result in a Class 1 misdemeanor (up to 1 year imprisonment, up to $2,000 fine) and civil liability for damages, costs, and attorney fees.

Sources and References

SourceType
2025 Senate Bill 164 | South Dakota Legislaturelegal
South Dakota Codified Laws - Chapter 12-26 (New Sections added by SB 164)legal
SD SB164 | 2025 | Regular Session - LegiScanlegal
Plain English

South Dakota's new Deepfakes Regulation, SB 164, aims to protect election integrity by prohibiting the undisclosed use of manipulated media in political campaigns. This law applies to any person who creates or shares deepfakes in South Dakota elections.

Specifically, the Act makes it illegal to spread a deepfake with the intent to harm a candidate within 90 days before an election. A "deepfake" is defined as an image, audio, or video created using artificial intelligence or other digital tools that makes a real person appear to say or do something they didn't, or fundamentally alters the understanding of actual events.

To comply, if you disseminate a deepfake, you must include a clear disclosure. For images and videos, this means superimposing the text: "This (image/video/audio) has been digitally generated or manipulated and is not an accurate representation of fact or actual events." This text must be easily readable and no smaller than other prominent text. For audio, the statement must be clearly spoken at both the beginning and end of the recording.

The law took effect on March 31, 2025. Violating this Act is a serious matter, carrying both criminal and civil consequences. - Criminal penalties include a Class 1 misdemeanor, which can mean up to one year in jail and/or a $2,000 fine. - Separately, civil penalties allow injured candidates or depicted individuals to sue for damages, legal costs, and attorney fees.

A key pitfall to remember is the strictness of the disclosure requirements. Even if you intend no harm, failing to meet the precise font size, placement, or audio timing for the disclosure could lead to liability. However, the law does exempt content that is clearly satire or parody, and media organizations not directly involved in creating or maliciously spreading the deepfake.

Plain-English rewrite by Regulations.ai — not legal advice. Verify against the official text.

What you must do — compliance checklist

0 / 9 marked complete

Plain-English obligations under United States - South Dakota - Deepfakes Regulation (SB 164). Not legal advice — verify against the official text before relying on it.

  1. #1CriticalWithin 90 days of an election

    Applies to: Any person disseminating deepfakes.

    prohibits any person from disseminating a deepfake with the intent to injure a candidate within 90 days of an election, unless the deepfake includes a specific, prominent disclosure.
  2. #2CriticalBefore dissemination

    Applies to: Any person disseminating deepfake images or videos.

    the required disclosure must state: 'This (image/video/audio) has been digitally generated or manipulated and is not an accurate representation of fact or actual events.'
  3. #3CriticalBefore dissemination

    Applies to: Any person disseminating deepfake images or videos.

    text must appear in a size that is easily readable... no smaller than the largest font size... superimposed over each deepfake
  4. #4CriticalBefore dissemination

    Applies to: Any person disseminating deepfake audio recordings.

    the statement must be read in a clearly spoken manner and in a pitch that is easily heard... at both the beginning and the end
  5. #5ImportantBefore dissemination

    Applies to: Any person creating or disseminating manipulated media.

    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.
  6. #6ImportantBefore dissemination

    Applies to: Any person disseminating deepfakes.

    with the intent to injure a candidate, disseminates a deepfake... knowing or reasonably knowing it is a deepfake
  7. #7ImportantBefore dissemination

    Applies to: Any person disseminating deepfakes.

    within 90 days of an election
  8. #8ImportantBefore dissemination

    Applies to: Any person disseminating deepfakes.

    exemptions from liability... for content that clearly constitutes satire or parody
  9. #9ImportantBefore dissemination

    Applies to: Broadcasters, newspapers, websites, and radio stations.

    exemptions from liability for broadcasters... provided they are not directly involved in the creation or intentional dissemination of the harmful deepfake

© Regulations.AI — created on 06-Jan-2026 using Gemini 2.5 Flash