Lawmakers Propose Right of Publicity Protections for Digital Replicas
Key highlights this week:
We’re already tracking 475 bills in the 2025 legislative session.
Lawmakers concerned about the unauthorized use of “digital replicas” have passed legislation establishing a property right in one’s own likeness.
An amended version of a high-risk artificial intelligence regulation bill passed out of committee in Virginia.
A proposal in Rhode Island would create strict liability for AI developers for injuries suffered by a user.
Last year, state lawmakers introduced hundreds of bills regarding sexual and political deepfakes, enacting several of them. Political deepfake bills sought to protect political candidates from misleading communications while sexual deepfake bills protected individuals from having embarrassing computer-generated images of them distributed. This year, concerns about the use of one’s image have spurred lawmakers in some states to consider legislation concerning digital replicas, which may impact advertising and other commercial media.
A digital replica differs from a deepfake in its purpose more so than in its creation. Deepfakes are AI replicas created with the intent to deceive, harass, or embarrass while a digital replica is an authorized recreation of an individual's likeness, generally for commercial purposes. For example, in 2023 Tom Hanks clarified that he had no involvement in an advertisement for a dental plan that featured an AI version of himself. This was a deepfake as the use of his likeness in the advertisement was likely meant to imply his endorsement, presumably to deceive consumers. However, if he had consented to the use of the image in the advertisement it would have been considered a digital replica. This is not to say that there cannot be unauthorized digital replicas. Many digital replica bills establish property rights in the replica, provisions for assigning rights, post-mortem rights, and civil liability for violations.
The digital replica bills being considered by lawmakers relate to the right of publicity, an intellectual property right that protects against the commercial misappropriation of a person’s name or likeness, including their image or voice. There is no federal law recognizing a right of publicity and while many states recognize a right of publicity, not all do. Of the states that do recognize the right of publicity, some have done so explicitly through statute, while others have recognized the right through case law.
So far this session, lawmakers in ten states have introduced legislation addressing digital replicas. Many of the bills introduced so far are in states that recognize the right of publicity, but these bills additionally clarify that an individual has a property right in their likeness which extends to recreations of their image and voice through the use of a digital replica.
Mississippi lawmakers have introduced their own version of Tennessee’s ELVIS Act (MS HB 768/SB 2796) which creates a property right in an individual’s name, image, voice, or likeness. Tennessee’s version of the ELVIS Act was signed into law last year and received praise from performers who had expressed concern about deepfakes of themselves being used to create unauthorized artistic works and or to promote products or services without their consent. Mississippi lawmakers are also considering another bill (MS HB 2220) that establishes similar property rights with specific provisions for digital replicas.
In Arkansas, lawmakers are acting quickly on AR HB 1071, which adds AI-generated images and audio to the definition of “likeness” in its right of publicity law. This bill was passed out of the House on Monday and is currently being considered by the Senate. Virginia lawmakers are also considering legislation (VA HB 2462) which creates a property right in the voice or likeness of a person in a digital replica for up to 100 years after their death and creates a private right of action for the production or unauthorized distribution of a digital replica. In the Senate, legislators considered legislation (VA SB 1421) that would have prohibited the unauthorized use of a deceased individual's voice or likeness, however, that bill failed to pass out of Committee on Wednesday.
Lawmakers in New Mexico have introduced legislation (NM HB 221) that establishes provisions for the commercial use of digital replicas, including regulations on how the digital replicas of minors can be used. Lawmakers in Maine are also focusing their attention on digital replicas of minors with legislation (ME LD 230) that explicitly prohibits using an AI-generated image of any actual person who is a minor in advertisements.
Not all digital replica legislation relates to the right of publicity. In New York, lawmakers are considering legislation (NY AB 606/SB 1228) that would require an advertisement to disclose if there is a synthetic performer. In Massachusetts, lawmakers have drafted two bills that establish a cause of action for the unauthorized use of a person’s likeness for advertising (MA HD 398) and require certain provisions to be in contracts governing the creation and use of digital replicas (MA HD 3986). Bills in Oklahoma (OK HB 2216) and South Carolina (SC HB 3404) do not specifically reference AI or digital replicas, but do prohibit infringing on the right of publicity “in any medium in any manner” and establish penalties for violations.
As the use of digital replicas becomes more prevalent in advertising and other commercial mediums, lawmakers will seek to give people rights over their own likeness.
Recent Developments
In the News
Chinese AI models: The tech world was rocked by the unveiling of a Chinese AI model called DeepSeek that can outperform U.S. models at a fraction of the cost using an open-source model. Later in the week, China-based Alibaba announced its model Qwen 2.5 outperformed DeepSeek-V3 and other AI systems, further spooking tech leaders and policymakers pushing for U.S. dominance in the technology.
Major Policy Action
Federal: The National Artificial Intelligence Advisory Committee approved a draft document that covers ten priority issues surrounding artificial intelligence, including awareness and literacy, workforce, education, health, government, governance, law enforcement, science, and small business. The draft includes recommendations to establish a model to evaluate, test and assess AI systems “that builds on the success of” FedRAM and increased funding for NIST’s AI Safety Institute.
Arkansas: The House passed a bill (AR HB 1071) to amend the Publicity Rights Protection Act to include images and voice generated through artificial intelligence. The original law was created to strengthen publicity rights for student-athletes.
Pennsylvania: The state Department of Education denied an application this week for a controversial cyber charter school that uses artificial intelligence as tutors with human “guides.” The department rejected the application for a number of reasons, including unrealistic projections for enrollment growth, whether the school could attain insurance, and its ability to support special education.
Virginia: A high-risk artificial intelligence system regulation bill (VA HB 2094) barely passed out of committee with a substituted version on Monday. Among many changes, the substituted version removes obligations for distributors of such systems and requires developers use a reasonable duty of care to protect consumers against reasonably foreseeable risks of algorithmic discrimination.
Notable Proposals
Georgia: Senator John Albers (R) introduced a bill (GA SB 37) that would create the Georgia Board for Artificial Intelligence to advise governmental entities through the publication of model comprehensive artificial intelligence system usage plans. Albers chaired the Senate Study Committee on Artificial Intelligence that issued a report in December that made policy recommendations on AI and is expected to introduce a regulatory AI bill along the lines of the Colorado law.
Maryland: Senator Katie Fry (D) is behind a number of AI-related bills, including a measure (MD SB 362) that would make it illegal to distribute a “forged digital likeness” that misrepresents the appearance, speech, or behavior of another and is likely to deceive. That measure is in response to a case where a former high school employee is accused of using AI to impersonate the principal to portray him making racist and antisemitic remarks.
New York: A new comprehensive regulatory bill (NY AB 3356) introduced by Assemblyman Clyde Vanel (D) would require registration and licensing of high-risk advanced artificial intelligence systems with the Department of State. The measure also requires notice of incidents to the Department and allows the agency to designate systems as posing a state or national security risk.
Rhode Island: A measure introduced by three Democrats (RI HB 5224) would impose strict liability for injuries to a non-user caused by a model. It would only apply to certain large models and allow an affirmative defense if the developer establishes that the model satisfied the standard of care applicable to humans who perform the same function.