United States - Arizona - Synthetic Media Regulation (SB 1359)

Overview

Arizona Senate Bill 1359, officially signed into law on May 29, 2024, and subsequently filed with the Secretary of State, represents a significant legislative effort by the State of Arizona to address the burgeoning challenges posed by artificial intelligence (AI) in the context of electoral integrity. This Act, now codified within Title 16, Chapter 7, Article 1 of the Arizona Revised Statutes, specifically targets the creation and distribution of synthetic media, commonly referred to as 'deepfakes,' during critical periods leading up to an election. The primary objective of this legislation is to ensure transparency in political communications by requiring explicit disclosures when AI-generated content is used to depict candidates or political parties in a deceptive and fraudulent manner. The legislative intent behind SB 1359 is to safeguard the democratic process from manipulative digital content that could mislead voters, thereby preserving public trust in election outcomes and the authenticity of political discourse. The rapid advancement and accessibility of AI technologies capable of generating highly realistic but fabricated images, audio, and video necessitate proactive regulatory measures to prevent their misuse in influencing public opinion and electoral results. By focusing on disclosure, Arizona aims to empower voters with the information needed to critically assess the authenticity of political messages they encounter, fostering a more informed electorate.

The scope of Arizona SB 1359 is carefully defined to address specific instances of synthetic media use that pose a direct threat to the integrity of elections. It applies to synthetic media messages created and distributed within ninety days before an election where a candidate for elected office will appear on the ballot. This temporal restriction underscores the legislature's focus on the period when electoral campaigning is most intense and voter decisions are being finalized. The Act specifically prohibits the creation and distribution of synthetic media that a person knows to be a 'deceptive and fraudulent deepfake' of a candidate or political party, unless it includes a clear and conspicuous disclosure indicating its AI-generated nature. This nuanced approach acknowledges the legitimate uses of AI in media while drawing a clear line against its deceptive application in political contexts. Furthermore, the legislation carves out important exemptions for media that constitutes satire or parody, as well as for interactive computer services, recognizing free speech considerations and the existing regulatory framework for online platforms. These exemptions are crucial for balancing the need for electoral integrity with protections for artistic expression and the operational realities of internet service providers.

Definitions

Arizona SB 1359 meticulously defines key terms to ensure clarity and precision in its application, thereby minimizing ambiguity in enforcement and compliance. Central to the Act are the definitions of 'Creator,' 'Deceptive and fraudulent deepfake,' and 'Synthetic media'. A 'Creator' is defined as any person that uses artificial intelligence or other digital technology to generate synthetic media. Crucially, this definition explicitly excludes the provider of any technology used in the creation of synthetic media. This distinction is vital, as it places the primary responsibility for disclosure on the individual or entity directly involved in generating the deepfake content, rather than on the developers or platforms that merely offer the underlying AI tools. This targeted definition ensures that the regulatory burden falls on those who actively produce and disseminate potentially misleading electoral content, aligning the Act's enforcement mechanisms with the actual points of content creation and distribution. The legislative intent behind this specific delineation is to prevent the chilling effect on technological innovation while directly addressing the actors responsible for the misuse of AI in elections.

The definition of 'Deceptive and fraudulent deepfake' is particularly critical, as it establishes the threshold for prohibited conduct under the Act. It means synthetic media that depicts a candidate or political party with the intent to injure the reputation of the candidate or political party or otherwise deceive a voter. Furthermore, such deepfake must either appear to a reasonable person to depict a real individual saying or doing something that did not actually occur in reality, or be intentionally calculated to mislead a reasonable person into concluding that a real individual said or did something that they did not say or do in reality. This dual-pronged definition emphasizes both the objective appearance of falsity and the subjective intent to deceive or harm, thereby setting a high bar for what constitutes a regulatable deepfake. It differentiates between genuine synthetic media and that which is maliciously crafted to manipulate public perception. 'Synthetic media' itself is broadly defined as an image, audio recording, or video. This comprehensive definition covers various forms of digital content that can be manipulated using AI, ensuring the Act's applicability across different media types and future technological advancements in AI-generated content. The precision in these definitions is paramount for effective legal interpretation and consistent application of the disclosure requirements.

Governance and Institutional Framework

The governance and institutional framework for Arizona SB 1359 is intrinsically linked to the existing electoral laws of the state, as the Act amends Title 16, Chapter 7, Article 1 of the Arizona Revised Statutes, which pertains to election communications. This integration means that the enforcement and oversight of deepfake regulations fall under the established mechanisms for managing election-related activities. The Arizona Secretary of State, along with county recorders and other election officials, traditionally plays a pivotal role in administering election laws, and their existing authority will likely extend to the provisions of this Act. While SB 1359 does not establish a new dedicated regulatory body for AI or synthetic media, it leverages the existing infrastructure and expertise within the state's election administration. This approach ensures that the new regulations are seamlessly incorporated into the broader legal framework governing political campaigns and voter information. The reliance on established institutions also implies that any disputes or enforcement actions related to deepfake violations would likely be handled through the existing legal and administrative channels designed for electoral matters, providing a degree of predictability and consistency in application.

The Act's placement within the Arizona Revised Statutes signifies a legislative intent to treat deceptive deepfakes as a serious matter within the realm of election integrity, akin to other forms of prohibited election communications. The enforcement mechanisms, including civil penalties and the availability of injunctive relief, are designed to be administered through the state's judicial system, allowing for legal recourse for aggrieved parties. This judicial oversight provides a critical check and balance, ensuring that enforcement actions are fair, proportionate, and consistent with due process. Furthermore, the Act's provisions implicitly rely on the public and political campaigns themselves to act as monitors, reporting potential violations to the appropriate authorities. The ability for a candidate whose appearance, action, or speech is depicted in a deceptive deepfake to seek injunctive or other equitable relief directly from the sponsor or creator of the media empowers individuals to protect their reputations and prevent the further dissemination of harmful content. This direct legal avenue serves as a vital component of the enforcement framework, offering a rapid response mechanism to address deepfake proliferation during sensitive election periods. The overall framework thus combines administrative oversight by election officials with judicial enforcement and individual legal remedies, creating a multi-layered approach to regulating synthetic media in Arizona elections.

Key Focus Areas

Arizona SB 1359 primarily focuses on ensuring transparency and accountability in political communications by addressing the specific threat of deceptive synthetic media during elections. A central tenet of the Act is the establishment of a critical 90-day window before an election, during which the creation and distribution of deceptive and fraudulent deepfakes are subject to strict disclosure requirements. This specific timeframe is strategically chosen to cover the most intensive period of political campaigning, when voters are actively engaged and susceptible to influence. Within this period, any person acting as a 'creator' is prohibited from creating and distributing a synthetic media message that they know to be a deceptive and fraudulent deepfake of a candidate or political party on the ballot, unless that message includes a clear and conspicuous disclosure. This disclosure must explicitly convey to a reasonable person that the media includes content generated by artificial intelligence. The emphasis on 'clear and conspicuous' disclosure is crucial, aiming to prevent subtle or easily overlooked disclaimers that might fail to inform the average viewer or listener effectively. The Act thus places a direct responsibility on the creators of such content to be transparent about its artificial nature, thereby empowering voters to make informed judgments about the authenticity of the information they consume.

Beyond the general requirement for visual disclosure, the Act also provides specific guidelines for synthetic media that consists solely of audio, where a visual disclosure is not feasible. In such cases, the disclosure must be read in a clearly spoken manner and in a pitch that can be easily heard by the average listener. This audio disclosure is mandated to be present at the beginning of the audio, at the end of the audio, and, if the audio is longer than two minutes, interspersed within the audio at intervals of not more than two minutes each. These detailed requirements underscore the legislature's commitment to ensuring that the disclosure is genuinely effective across different media formats, preventing circumvention of the transparency mandate. Furthermore, the Act includes important exemptions to its provisions, recognizing that not all synthetic media warrants regulation. Specifically, media that constitutes satire or parody is exempt, protecting artistic and comedic expression. Additionally, an interactive computer service, as defined in 47 United States Code section 230, is also exempt. This exemption aligns with federal law governing online platforms, acknowledging their role as intermediaries rather than primary content creators in many instances. Another exemption covers bona fide newscasts, news interviews, or news documentaries, provided they clearly acknowledge questions about the authenticity of the materially deceptive media. These exemptions are vital for balancing the Act's objectives of electoral integrity with fundamental rights such as freedom of speech and the practical realities of digital media distribution.

Implementation Framework

The implementation framework for Arizona SB 1359 relies heavily on the self-regulatory capacity of political campaigns and media creators, coupled with the enforcement powers of the state's legal system. The core of the implementation lies in the mandatory disclosure requirement: any person acting as a creator who distributes a deceptive and fraudulent deepfake within 90 days before an election must include a clear and conspicuous disclosure that the content is AI-generated. This places the onus directly on the individual or entity producing and disseminating the synthetic media to adhere to the transparency standards. For visual media, this implies a visible text overlay or similar prominent indicator. For audio-only content, the Act specifies precise requirements for spoken disclosures, including placement at the beginning, end, and at regular intervals for longer recordings. These detailed specifications aim to provide clear guidance to creators on how to comply, thereby reducing ambiguity and facilitating consistent application of the law. The practical implication is that political campaigns and individuals engaging in political communication must exercise due diligence to ascertain whether their media content falls under the definition of a 'deceptive and fraudulent deepfake' and, if so, ensure proper disclosure.

The implementation framework also includes provisions for enforcement and redress, which are crucial for the Act's effectiveness. A person who fails to make the required disclosure is liable for civil penalties. This financial deterrent is intended to encourage compliance and penalize non-adherence. Furthermore, the Act empowers candidates whose appearance, action, or speech is depicted in a deceptive and fraudulent deepfake to seek injunctive or other equitable relief from the sponsor or creator of the media. This provision offers a direct legal avenue for aggrieved parties to halt the further distribution of non-compliant deepfakes, providing a timely remedy during the sensitive election period. The ability to seek injunctive relief is particularly important in electoral contexts, where the rapid spread of misinformation can have immediate and irreversible impacts. The framework's success will depend on a combination of voluntary compliance by creators, active monitoring by campaigns and the public, and robust enforcement by state authorities through civil penalties and judicial remedies. The exemptions for satire, parody, and interactive computer services, as well as bona fide news, are also integral to the implementation, as they define the boundaries of the Act's reach and prevent overreach into protected forms of expression or established media practices.

Monitoring and Evaluation

The monitoring and evaluation of Arizona SB 1359, while not explicitly detailing a dedicated state agency or program for this purpose, will largely occur through a combination of public vigilance, the actions of political campaigns, and the existing enforcement mechanisms within Arizona's election laws. The Act implicitly relies on candidates, political parties, and the general public to identify and report instances of non-compliant synthetic media. Given the high stakes of elections, it is anticipated that opposing campaigns and vigilant citizens will actively monitor political communications for deceptive deepfakes that lack the required disclosures. This decentralized monitoring approach leverages the collective interest in maintaining fair and transparent elections. Any reported violations would then trigger the enforcement provisions of the Act, leading to investigations by relevant state authorities, potentially including the Arizona Attorney General's office or other bodies responsible for enforcing election laws. The effectiveness of this monitoring system will depend significantly on public awareness of the Act's provisions and the ease with which potential violations can be reported and acted upon. Educational initiatives for voters and campaigns about the dangers of deepfakes and the requirements of SB 1359 could enhance this monitoring capacity.

Furthermore, the judicial process itself serves as a critical mechanism for evaluating the Act's application and impact. When candidates seek injunctive relief or when civil penalties are pursued, courts will interpret the definitions of 'deceptive and fraudulent deepfake' and 'clear and conspicuous disclosure,' thereby shaping the practical contours of the law. The outcomes of these legal challenges will provide valuable precedents and insights into how effectively the Act addresses its intended purpose. Over time, an analysis of the number of reported violations, the success rate of enforcement actions, and the types of synthetic media targeted could inform future legislative adjustments. While the Act does not mandate a formal review or sunset clause, the ongoing evolution of AI technology and its potential misuse in elections will likely necessitate periodic legislative re-evaluation. The background information provided in the Senate Fact Sheet, referencing President Biden's Executive Order 13960 on AI, suggests a broader national and potentially international context for regulating AI-generated content. This wider conversation may also influence future assessments of Arizona's deepfake legislation, as states and federal entities continue to grapple with the societal implications of advanced AI. Therefore, while no formal evaluation framework is outlined, the combination of public oversight, judicial review, and the dynamic nature of AI technology will collectively contribute to the ongoing monitoring and de facto evaluation of SB 1359's efficacy.

Penalties, Liability, and Appeals

Arizona SB 1359 establishes clear consequences for non-compliance with its disclosure requirements, primarily through civil penalties and the provision for injunctive relief. A person who fails to make the clear and conspicuous disclosure required by the Act for a deceptive and fraudulent deepfake is liable for the civil penalty prescribed by section 16-937, subsection B, of the Arizona Revised Statutes. This specific reference to an existing statute integrates the new deepfake regulations into the established framework for election law violations, providing a pre-defined penalty structure. Section 16-937, subsection B, generally pertains to civil penalties for campaign finance violations, indicating that the legislature views the undisclosed distribution of deceptive deepfakes with a similar level of seriousness as other breaches of electoral transparency. The penalty is applied for each day that the deceptive and fraudulent deepfake is distributed without the required disclosure, underscoring the continuous nature of the violation and providing a strong incentive for prompt removal or correction of non-compliant content. This daily accrual mechanism is designed to deter prolonged dissemination of misleading AI-generated political messages, ensuring that violations are addressed swiftly during sensitive election periods. The civil nature of these penalties means that enforcement actions would typically involve monetary fines rather than criminal charges.

Beyond civil penalties, the Act provides a crucial mechanism for immediate redress for affected candidates. A candidate whose appearance, action, or speech is depicted through the use of a deceptive and fraudulent deepfake in violation of this section may seek injunctive or other equitable relief from the sponsor or the creator of the media. This provision empowers candidates to directly petition a court to prohibit the publication or further distribution of the deceptive deepfake. The availability of injunctive relief is particularly significant in the context of elections, where the rapid dissemination of false information can cause irreparable harm to a candidate's reputation and campaign prospects in a short timeframe. Equitable relief allows courts to issue orders that compel specific actions, such as removing content, or prohibit certain activities, such as continued distribution, offering a more immediate and targeted remedy than monetary damages alone. It is important to note that earlier versions of the bill may have included criminal penalties, but the final enacted version focuses on civil penalties and injunctive relief. This shift reflects a legislative decision to address deepfake misuse through civil enforcement mechanisms, which are often more agile and appropriate for addressing content-related disputes in a political context. The appeals process for any civil penalties or injunctive orders would follow standard Arizona judicial procedures, allowing for due process and review by higher courts.

Relationship to Other Instruments

Arizona SB 1359 is strategically integrated into the existing legal framework governing elections within the state, specifically by amending Title 16, Chapter 7, Article 1 of the Arizona Revised Statutes. This integration means that the deepfake regulations do not stand in isolation but rather become an organic part of the comprehensive body of law that dictates how political campaigns are conducted, how election communications are managed, and how electoral integrity is maintained. By amending an existing chapter, the Act leverages established legal definitions, enforcement procedures, and judicial interpretations that apply to other aspects of election law. This approach provides a degree of legal consistency and avoids the creation of an entirely new and potentially conflicting regulatory regime. The Act's reliance on existing civil penalty structures, such as those referenced in section 16-937, subsection B, further demonstrates this interconnectedness, ensuring that the consequences for deepfake violations are aligned with other forms of electoral misconduct. This relationship to existing instruments streamlines the implementation process, as election officials and legal practitioners can draw upon familiar legal precedents and administrative practices when addressing issues related to synthetic media.

Furthermore, Arizona SB 1359 also explicitly acknowledges and interacts with federal law, particularly concerning the exemption for interactive computer services. The Act states that its provisions do not apply to an interactive computer service as defined in 47 United States Code section 230. This reference is highly significant, as Section 230 of the Communications Decency Act provides broad immunity to online platforms for content posted by third parties. By incorporating this federal carve-out, Arizona's deepfake legislation avoids potential conflicts with federal internet law and recognizes the distinct legal status of platforms that host user-generated content versus the creators of that content. This exemption is crucial for maintaining the operational viability of social media platforms, search engines, and other online services, preventing them from being held liable for every deepfake disseminated by their users. The Act's Senate Fact Sheet also indirectly references a broader federal context by mentioning President Biden's Executive Order 13960, which aims to promote safe AI development and prevent the spread of deepfakes. While a state law, this indicates an awareness of national efforts to address AI governance and the challenge of synthetic media, suggesting a potential for future alignment or coordination with federal guidelines as they evolve. This interplay with both state election statutes and federal internet law demonstrates a considered approach to regulating AI in a complex legal landscape.

International Alignment

While Arizona SB 1359 is a state-level regulation within the United States, its existence and focus on synthetic media in elections align with a growing international concern regarding the impact of artificial intelligence on democratic processes. The global proliferation of AI-generated content, particularly deepfakes, has prompted legislative and policy discussions in numerous countries and international bodies. Many nations are grappling with how to balance freedom of expression with the need to combat misinformation and protect electoral integrity. Arizona's Act, by mandating disclosure for deceptive deepfakes in political campaigns, contributes to this global trend of increasing transparency requirements for AI-generated content, especially in sensitive areas like elections. The underlying principles of the Act—transparency, accountability, and the prevention of voter deception—resonate with broader international efforts to establish ethical guidelines and regulatory frameworks for AI. Although there is no direct formal alignment with specific international treaties or supranational laws, the spirit of the Arizona law reflects a shared global recognition of the risks posed by unchecked AI in public discourse.

The Senate Fact Sheet for SB 1359 even references President Biden's Executive Order 13960, which aims to promote the safe and trustworthy development and use of AI, including directives to prevent the spread of deepfakes. This mention highlights that even state-level initiatives are often informed by or contribute to a broader national strategy, which itself may be influenced by international dialogues and best practices in AI governance. Many international organizations, such as the OECD, UNESCO, and the European Union, have been actively developing recommendations, guidelines, and regulations for AI, often emphasizing principles like transparency, explainability, and human oversight. The EU's AI Act, for instance, includes provisions for transparency requirements for certain AI systems, including those generating synthetic content. While Arizona's law is distinct, its focus on mandating disclosure for AI-generated political content mirrors the transparency objectives seen in these international discussions. The Act demonstrates a localized response to a global challenge, contributing to the evolving patchwork of regulations designed to mitigate the risks of AI while harnessing its potential responsibly. As AI technology continues to advance, it is likely that state, national, and international efforts will continue to converge on common principles for managing its societal impact, particularly in critical domains like elections.

Implementation Timeline

Compliance Checklist

Sources and References