New Jersey Deepfake Laws: AI Images, Voice Cloning & Penalties (2026)

New Jersey enacted one of the country's broadest deepfake laws when Governor Murphy signed A3540 (P.L. 2025, c.40) on April 2, 2025. The law, which took effect immediately, creates both criminal and civil liability for producing or disclosing deceptive audio or visual media used to further a crime. It covers intimate images, election manipulation, voice cloning, and child endangerment within a single comprehensive statute codified at N.J. Stat. Ann. § 2C:21-17.7 and § 2C:21-17.8.

Is It Illegal to Make a Deepfake of Someone in New Jersey?

Yes, under the right circumstances, and New Jersey's law is broader than most. P.L. 2025, c.40 does not require that a deepfake be sexual in nature to trigger criminal liability. The law targets deepfakes used to further any crime, a sweeping predicate-offense model that covers intimate images, election interference, harassment, threats, and child endangerment under a single statute.

The law defines "deceptive audio or visual media" as any video, motion picture, sound recording, electronic image, photograph, technological representation of speech, or document forgery that appears to a reasonable person to realistically depict speech, conduct, or writing the depicted person did not actually engage in, and whose production was substantially dependent on technical means rather than a human imitator. That definition captures AI-generated content, deepfake video, voice clones, and synthetic text documents.

Three buckets summarize what New Jersey law covers:

• Sexual and intimate deepfakes: Creating a deepfake with intent to harass, cyber-harass, or commit child endangerment is a third-degree crime. Civil liability runs parallel.
• Election and political deepfakes: Deepfakes used to make threats or exert improper influence in official and political matters are explicitly listed as predicate offenses.
• Voice cloning and digital likeness: The statute covers sound recordings and technological representations of speech, meaning AI voice clones are covered when used to further a crime. New Jersey has no separate right-of-publicity statute covering voice, but the 2025 deepfake law provides meaningful criminal and civil protection.

The statute also reaches people who never touch the software. Subsection (c)(1) criminalizes soliciting a deepfake at the same third-degree level as making one, and "solicit" is defined broadly: it covers offering to create or advertising the ability to create deceptive media, which is the business model of the so-called nudify apps, and it separately covers simply requesting that one be created, "whether or not for monetary remuneration or compensation." Commissioning a deepfake from someone else is therefore its own third-degree exposure, not a way around the law.

Exemptions protect criticism, comment, satire, parody, news reporting, teaching, scholarship, and research, along with content a reasonable viewer or listener would not believe authentically depicts speech or conduct. A separate exemption covers interactive computer service providers, cloud services providers, and commercial developers of artificial intelligence technology, but the condition is a Section 230 status test, not a state-of-mind test: the exemption applies where the provider "is not deemed to be a publisher, speaker, or information content provider pursuant to 47 U.S.C. s.230." Whether a given distribution was inadvertent is not what the statute asks.

Sexual and Intimate Deepfakes

New Jersey's deepfake law directly addresses intimate-image abuse. Harassment and cyber-harassment are listed predicate offenses under § 2C:21-17.8, meaning a person who uses AI to fabricate explicit images of another person and sends or posts them commits a third-degree crime, regardless of any prior relationship with the victim.

The penalty structure has three tiers, and two of them are third-degree. Creating or generating a deepfake for criminal purposes is a third-degree crime carrying 3 to 5 years in New Jersey State Prison and a fine up to $30,000. Soliciting, disclosing, or using a deepfake for the purpose of furthering a crime, or knowing another person will use it that way, is a separate third-degree crime under subsection (c)(1) carrying the same 3 to 5 years. Only the residual tier is lighter: knowingly or recklessly disclosing a deepfake that was created in violation of the law, without acting for that criminal purpose, is a fourth-degree crime punishable by up to 18 months.

That distinction matters to anyone who forwards rather than creates. A person who passes along a fabricated explicit image in order to harass the person depicted falls under the third-degree tier, not the 18-month one. The statute also blocks the usual sentencing discounts: a conviction under the creation subsection does not merge with a conviction under the disclosure subsection, neither one merges with the underlying offense it was used to further, and the court must impose consecutive sentences.

The law was partly inspired by Francesca Mani, a student at Westfield High School who was victimized by peer-created AI intimate images. Her public advocacy directly influenced Governor Murphy's prioritization of the legislation in the 2025 session.

For victims under 18, N.J. Stat. Ann. § 2C:24-4 defines "reproduction" to include computer-generated images, covering AI-generated child sexual abuse material as child endangerment under existing law. P.L. 2025, c.40 reinforces this by listing endangering the welfare of children as a named predicate offense. Federal law under 18 U.S.C. 2256 (the PROTECT Act) independently covers AI-generated CSAM indistinguishable from a real minor, providing a parallel federal floor regardless of any state gap.

On the civil side, victims may sue the perpetrator for actual damages or liquidated damages of $1,000 per violation (whichever is greater), plus punitive damages and attorney's fees, without first obtaining a criminal conviction. This standalone civil cause of action can be filed alongside, before, or after any criminal proceeding and supplements any common law claims for defamation, invasion of privacy, false light, or misappropriation of identity.

Election and Political Deepfakes

P.L. 2025, c.40 specifically covers deepfakes used for "threats or improper influence in official and political matters" as a predicate offense. A person who fabricates a deepfake of a political candidate, election official, or public figure to threaten, coerce, or manipulate an election outcome commits a third-degree crime under the statute.

New Jersey is among roughly 28 states with some form of election deepfake protection. The predicate-offense model in § 2C:21-17.8 differs from disclosure-only approaches used in some other states. New Jersey does not merely require a disclosure label on AI-generated political ads; it criminalizes using deepfakes as a tool of election-related threats or improper influence.

The First Amendment dimension deserves note. In August 2025, a federal court in Kohls v. Bonta struck down California's election deepfake statute (AB 2839) in its entirety and permanently enjoined its enforcement on First Amendment grounds, illustrating that broad restrictions on political speech carry constitutional risk. New Jersey's threat-and-improper-influence framing targets conduct rather than content, which is a more defensible approach, but litigation in this space remains ongoing and the law's contours will be shaped by future cases.

AI Voice Cloning and Digital Likeness

New Jersey's 2025 deepfake law covers "sound recording" and "technological representation of speech" within its definition of deceptive audio or visual media. An AI-generated voice clone used to produce content that falsely depicts what someone said, and that is used to further a crime, falls squarely within § 2C:21-17.8.

For comparison, Tennessee's ELVIS Act (Tenn. Code Ann. 47-25-1101, eff. July 1, 2024) is the national archetype for AI voice-clone legislation, extending the right of publicity expressly to voice simulations created by AI. New Jersey does not have a comparable standalone right-of-publicity statute protecting voice or likeness for commercial use. The 2025 deepfake law fills part of that gap on the criminal and civil side, but only when a voice clone is used to further a crime. A non-criminal commercial misuse of someone's AI-simulated voice, such as an unauthorized advertising campaign, would not trigger P.L. 2025, c.40 and would instead rely on common law claims (misappropriation of identity, unfair competition) without a robust statutory foundation.

For broader New Jersey AI regulation beyond deepfakes, including the state's AI governance bills addressing algorithmic transparency and automated decision-making, see New Jersey AI Laws and Regulation.

Federal Law That Applies in New Jersey

Several federal laws layer onto New Jersey's state protections and fill gaps where state coverage is absent.

The TAKE IT DOWN Act (Public Law 119-12, signed May 19, 2025) is the first federal law specifically targeting nonconsensual intimate deepfakes. It makes it a federal crime to knowingly publish intimate visual depictions of adults or minors, including AI-generated "digital forgeries," without consent. Penalties reach up to 2 years in federal prison, or 3 years when a minor is depicted. Platforms must remove flagged content within 48 hours of victim notice, with compliance enforcement by the FTC. The 48-hour removal obligation took effect May 19, 2026. New Jersey residents can use this federal avenue in parallel with the state law.

The FCC's AI-robocall ruling (FCC 24-17, February 2024) clarifies that AI-generated voices in robocalls are "artificial voices" under the Telephone Consumer Protection Act, making AI voice-clone calls to phones without prior express consent illegal under federal law. This ruling was triggered by the 2024 New Hampshire fake-Biden primary robocall and applies nationwide, including in New Jersey.

The FTC's Impersonation Rule (16 CFR Part 461, effective April 1, 2024) prohibits deceptive impersonation of government entities and businesses via AI voice cloning and other means. The individual-impersonation extension remains a proposed rule, not yet final.

Two federal proposals remain pending and are not law. The DEFIANCE Act (S.1837, 119th Congress) would create a federal civil cause of action for deepfake sexual abuse victims with liquidated damages of $150,000 (or $250,000 if the conduct involved actual or attempted sexual assault, stalking, or harassment). S.1837 passed the Senate by unanimous consent on January 13, 2026 and is now pending in the House. The NO FAKES Act (S.4591, 119th Congress) would establish a federal voice-and-likeness right against unauthorized AI digital replicas and has cleared the Senate Judiciary Committee and awaits a Senate floor vote (as of August 2026). Neither bill has become law as of mid-2026.

What Victims Can Do

Victims of deepfakes in New Jersey have multiple overlapping avenues for relief.

For criminal enforcement, report to local law enforcement or the New Jersey Division of Criminal Justice. Soliciting, disclosing, or using a deepfake to harass or cyber-harass someone is a third-degree crime; the fourth-degree tier applies only to knowing or reckless disclosure of an unlawfully created deepfake by a person not acting for that purpose. There is no separate, lighter penalty for forwarding rather than posting: the statute defines "disclose" to mean to "sell, give, provide, lend, trade, mail, deliver, transfer, publish, distribute, circulate, disseminate, present, exhibit, advertise, offer, share, or make externally available," so distribution is disclosure. The dividing line is purpose, not the manner of transmission. Referrals can also be made to the FBI's Internet Crime Complaint Center (IC3) at ic3.gov for federal TAKE IT DOWN Act violations.

For platform removal, the TAKE IT DOWN Act's 48-hour rule applies nationwide as of May 2026. Major platforms must process victim removal requests within 48 hours of notice. The National Center for Missing and Exploited Children (NCMEC) operates a reporting tool tied to the TAKE IT DOWN program.

For civil remedies, P.L. 2025, c.40 creates a standalone civil cause of action against any person who violates the statute. Victims may recover actual damages with a $1,000-per-violation liquidated floor, punitive damages, and attorney's fees. No criminal conviction is required first. Victims may also assert common law claims for defamation, invasion of privacy, false light, or misappropriation of identity in the same civil action.

For election-related deepfakes, victims and targeted candidates should contact the New Jersey Election Law Enforcement Commission (ELEC) and may seek injunctive relief in Superior Court to stop dissemination of the content.

Penalties at a Glance

Disclaimer: This page provides general legal information about New Jersey deepfake laws and is not legal advice. This area of law is changing rapidly, with new legislation and federal proposals moving on short notice. If you have been harmed by a deepfake or face a deepfake-related legal matter, consult a licensed New Jersey attorney.

More New Jersey Laws

• New Jersey AI Meeting Recording Laws
• New Jersey Data Privacy Laws
• New Jersey Recording Laws
• New Jersey Sexting Laws