Ahead of this year’s U.S. presidential election, Utah’s governor signed SB 131 (the “Bill”) into law on March 13, 2024. SB 131 amends the state’s Information Technology Act to require certain disclosures related to the use of AI and synthetic media in political advertising. Together with Utah’s recently passed Artificial Intelligence Policy Act, the Bill signals a growing effort among U.S. states to regulate AI technologies.
The bill takes effect on May 1, 2024, leaving little time for companies to come into compliance. The risk of non-compliance can be significant—the Bill provides for a private right of action and penalty of $1,000 per violation, making this an attractive target for plaintiff’s lawyers.
Scope
The Bill defines AI and generative AI in reasonably specific terms that avoid encompassing machine-assisted decisions or content that are not typically considered to be AI.
- “Artificial intelligence” means “a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments.”
- “Generative artificial intelligence” means “artificial intelligence technology that is capable of creating content such as text, audio, image, or video based on patterns learned from large volumes of data rather than being explicitly programmed with rules.
The Bill applies to audio or visual communications that contain “synthetic media” (meaning audio, image, or video content that are substantially produced by generative AI) and are:
- paid for by a candidate campaign committee, political action committee, political issues committee, political party, or a person using a contribution; and
- intended to influence voting for or against a candidate or ballot proposition in an election or primary in Utah.
Disclosure Obligations
Political ads containing synthetic media must explicitly disclose that the content was substantially produced by generative AI via a disclosure tailored to the particular style of media.
For audio-only political ads, the words “Contains content generated by AI” must be included audibly at the beginning and end of the communication. For visual political ads like videos and photos, the following rules apply. The required disclosures must be in “legible writing” and must be “displayed” throughout the portion of the communication containing synthetic media.
- Images: If an image contains synthetic media that is only visual, not audial, the disclosure must read: “This image generated by AI.”
- Video
- If the synthetic media is only visual, not audial, the disclosure must read: “This video content generated by AI.”
- If the synthetic media is only audial, not visual, the disclosure must read: “This audio content generated by AI.”
- Videos and Other Audio-visual Content: If a communication includes both synthetic audio and synthetic visual media, the disclosure must read: “This content generated by AI.”
Individuals who create or pay for political ads and publish such ads online must ensure that any such ads that are viewable, audible, or accessible in Utah include an embedded, tamper-evident “digital content provenance.” The provenance for political ads must disclose (1) the initial author and creator of the content; (2) any subsequent entities edited, altered, or otherwise modified; and (3) any use of generative AI in generating or modifying the substantive content.
As noted at the beginning of this post, the Bill is one of many emerging AI regulations in the U.S. Those creating or sponsoring political ads with generative AI will need to consider both the context-specific requirements of the Bill and the broader, developing requirements applicable to generative AI in Utah and other U.S. states.
