House Bill 375 would create a new Chapter 170 in the North Carolina General Statutes, titled the “Artificial Intelligence and Synthetic Media Act.” The bill is aimed at regulating deepfakes, synthetic media, and other AI-generated content, especially in political campaigns and in contexts involving minors and intimate images. It defines a wide range of terms related to artificial intelligence, synthetic media, digital impersonation, fabricated intimate images, and generated child pornography, and then sets out separate rules for elections, pornography-related offenses, civil remedies, and disclosure requirements for AI use.
In the election context, the bill would prohibit the creation or distribution of deceptive deepfakes of candidates or political parties within 90 days of an election unless the content includes prominent disclosures. It also requires certain online political ads to carry tamper-evident digital provenance information identifying the creator, editors, and any AI use. Outside elections, the bill would require disclosure when materially deceptive media is distributed, with exceptions for satire, parody, bona fide news reporting, and certain internet services. It also creates criminal penalties and civil remedies for violations, including expedited court procedures and injunctive relief.
The bill also addresses harmful AI-generated sexual content. It would make it a Class A felony to intentionally create or knowingly possess or view generated child pornography, and it would create a separate offense for knowingly disclosing fabricated intimate images without consent and with knowledge that harm is likely. Victims could sue for damages, statutory damages, punitive damages, attorneys’ fees, and injunctive relief. The bill further includes special handling rules for such evidence in criminal proceedings, including limits on copying, sealing requirements, and destruction procedures.
Beyond those targeted harms, the bill would require disclosure when generative AI is used in interactions with consumers, especially in regulated occupations such as licensed or certified professions. It also states that AI-assisted conduct does not avoid criminal liability under other statutes, and that the chapter should be construed broadly to protect voters and minors while preserving compliance with federal civil rights and disability laws and First Amendment protections. The act would take effect December 1, 2025, and apply prospectively to conduct on or after that date.
The overall sentiment reflected in the available context is limited because there were no recorded committee transcripts or votes, but the bill’s structure suggests a protective, anti-abuse purpose rather than a deregulatory one. The main points of contention likely center on free-speech concerns, the scope of liability for media outlets and online platforms, the breadth of the definitions of deceptive media, and whether the criminal penalties and disclosure mandates could be difficult to administer or overbroad in practice. The bill includes several exceptions and First Amendment references, indicating an attempt to address those concerns while still imposing significant restrictions on harmful AI-generated content.
HB375 would add a new chapter to the General Statutes regulating artificial intelligence and synthetic media, creating new criminal offenses, civil causes of action, disclosure obligations, and evidentiary rules. It would affect election law, criminal law, privacy and image-rights law, and professional disclosure requirements, while also imposing duties on creators, sponsors, publishers, and some regulated occupations. The bill would also create new felony and misdemeanor penalties and authorize injunctive relief, damages, and expedited court proceedings in certain cases.
No committee debate or vote history is available in the provided context, so there is no documented floor or committee sentiment to measure directly. Based on the bill text, the measure appears motivated by concerns about election integrity, child exploitation, and nonconsensual intimate imagery, with an emphasis on consumer and voter protection. At the same time, the inclusion of exceptions for satire, parody, news reporting, and First Amendment-protected speech suggests the bill anticipates legal and policy concerns about overreach.
Likely areas of contention include whether the bill’s deepfake and materially deceptive media definitions are too broad, whether the disclosure requirements burden political speech and journalism, and whether the criminal penalties are proportionate. Media organizations, online platforms, and AI developers may object to compliance burdens and potential liability, while civil liberties advocates may focus on free-speech and vagueness concerns. Supporters are likely to emphasize voter protection, anti-fraud enforcement, and stronger remedies for victims of AI-generated sexual abuse and impersonation.