Bill
S 2544
Creates criminal and civil liability for creating, soliciting, using, or disclosing deceptive deepfake media that facilitates crimes or harms, with specific penalties and remedies.
Status: Read twice; referred to Judiciary (most recent action 2025-07-30). Reported by Senate Law & Public Safety Committee (with amendments) and by Senate Budget & Appropriations Committee as a committee substitute. Hearing(s) scheduled (July 22, 2025). Introduced: 2024 (committee activity 2024–2025). Sponsors (listed in materials): Sen. Paul D. Moriarty; Sen. Kristin M. Corrado; Sen. Zellnor Myrie (and other cosponsors noted).
Purpose
- Establishes criminal and civil liability for the unlawful creation, solicitation, use, or disclosure of “deceptive audio or visual media” (commonly called “deepfakes”), particularly where such material is created or used to facilitate other crimes or to cause harm.
Key definitions (selected)
- Deceptive audio or visual media: video, film, sound recording, electronic image, photograph, technological representation of speech/conduct, or forgery/facsimile of a document that appears to a reasonable person to realistically depict speech, conduct, or writing of a person who did not in fact engage in it — and whose production was substantially dependent on technical means (e.g., generative AI), not merely human impersonation.
- Disclose / Solicit: broadly defined to cover selling, sharing, publishing, requesting, advertising the creation of, or otherwise making available such works.
- Victim: anyone who suffers personal, physical, or psychological injury or property loss as a result; includes certain family members in a death.
Criminal provisions
- Creates multiple third‑degree offenses for natural persons (i.e., individuals):
- Generating/causing generation of deceptive media with the intent it be used as part of a plan/course of conduct to commit crimes (enumerated categories include sexual offenses, bias crimes, theft/forgery/fraud, perjury/obstruction, harassment, cyber‑harassment, hazing, endangering children, threats/improper political influence, false public alarms, etc.).
- Soliciting, using, or disclosing deceptive media as part of a plan to commit such crimes.
- Disclosing deceptive media that the person knows or reasonably should know was created in violation of the statute.
- Penalties: third‑degree crime — ordinarily 3 to 5 years imprisonment and up to $15,000 fine; the bill authorizes courts to impose fines up to $30,000 (and requires non‑merging of these convictions with underlying offenses; consecutive sentences may be imposed).
- Evidentiary inference: a factfinder may infer lack of license/privilege when a work was created using commercially/publicly available generative AI in violation of that system’s terms of service.
Civil remedies
- Victims may sue in Superior Court (criminal conviction not required).
- Remedies include actual damages (not less than $1,000 per violation), punitive damages (for willful/reckless conduct), reasonable attorney’s fees and costs, and equitable relief. Civil remedies are cumulative of other available actions (defamation, invasion of privacy, etc.).
- A final State criminal judgment precluding denial of the same conduct in civil actions (estoppel).
Exceptions and limits
- First Amendment / fair‑use style exceptions: content clearly identified as deceptive (or clearly identified as possibly deceptive where nature is unknown) disclosed for criticism, comment, satire, parody, news reporting, teaching, scholarship, or research is treated as fair use and not subject to liability.
- Does not alter or negate rights/immunities of interactive computer service providers under 47 U.S.C. §230. The bill generally exempts interactive computer service providers, cloud providers, and AI developers/providers (so long as they are not treated as publishers), and contains tailored exceptions for news media, broadcasters, advertisers, and advertising platforms where certain identification or disclaimers are provided.
- Exempts disclosures in connection with law enforcement investigations, court orders, or certain cybersecurity/fraud investigations.
Procedural / fiscal impact
- Office of Legislative Services (multiple fiscal notes): the bill may produce indeterminate increases in State and local expenditures and revenues. Potential cost drivers: additional prosecutions (county prosecutors, Dept. of Law & Public Safety), additional judicial caseload, increased public‑defender representation, and possible incarceration and parole supervision (though first‑time third‑degree offenders are generally presumed for non‑incarceration). If incarceration occurs, OLS cites FY2023 average annual Department of Corrections cost ~ $75,574 per inmate. Fines and court fees could raise indeterminate revenue, though collection historically limited.
Who is affected
- Natural persons who create, solicit, use, or disclose deepfakes for unlawful ends.
- Victims of deepfake harms (including persons depicted and others harmed by related conduct).
- Law enforcement, prosecutors, the courts, public defenders, corrections and parole agencies.
- Interactive computer service providers, AI developers, news organizations and advertisers are carved out with specific exceptions and limits.
Notes / context
- The bill’s text evolved through committee substitutes to broaden enumerated offenses, add civil remedies, clarify fair‑use and §230 exceptions, and to add inferences tied to terms‑of‑service violations for generative AI systems. The introduced version also carried an alternative short title in some drafts (“GUARD Act”), as reflected in the bill text.
Compiled from official sources — confirm details with the bill’s official record.
Sign in to ask a question.
