We saw in a former post how S5959D /A5605-C became law, creating a post-mortem right of publicity in New York.
The law, which became effective on May 29, 2021, also created a “private right of action for unlawful dissemination or publication of a sexually explicit depiction of an individual,” codified at CVR § 52-C. This new Section of New York Privacy Law thus addresses the issue of pornographic “deepfakes.” California already enacted a similar law in 2019, Cal. Civ. Code § 1708.86.
“Deepfake” is a word referring to “deep learning” and “fake.” Deep learning refers to Artificial Intelligence (AI) technology. AI can, of course, be used in many ways, but can also be used to create deepfakes, which are lifelike digital representations of an individual. The technology needed to create a deepfake is now affordable and easy to use, placing it within the reach of average users.
Deepfakes are manipulated content and for that reason alone may raise concern. A report of the University College London published in 2020 by Crime Science ranked “deepfakes” as the most serious AI crime. This technology is particularly apt to be used for “revenge porn” purposes, which is a misdemeanor under New York law, or to fuel a campaign of hate against an individual, most likely a woman.
Such was the case in 2018 when a pornographic deepfake of an Indian investigative journalist was widely shared on social media and by text messages, disclosing her name, address, phone number and the words “I am available,” in an attempt to intimidate and discriminate against her. More recently, the mother of a cheerleader was arrested last March for having created deepfakes of three of her daughter’s cheerleading team members for the purpose of harassing them. The deepfakes showed the three teenagers in the nude, drinking alcohol or vaping.
What is a pornographic deepfake?
A pornographic deepfake is created when the face of an individual is morphed into a body engaged in a sexual activity, making it appear as if the person is indeed engaged in this activity. This is defined by Section 52-C-1-a as:
“an individual who appears, as a result of digitization, to be giving a performance they did not actually perform or to be performing in a performance that was actually performed by the depicted individual but was subsequently altered.”
The meaning of “digitization” is “to realistically depict the nude body parts of another human being as the nude body parts of the depicted individual, computer-generated nude body parts as the nude body parts of the depicted individual engaging in sexual conduct… in which the depicted individual did not engage.”
Possible Defenses
The defendant must have known or reasonably should have known that plaintiff did not consent to the “creation, disclosure, dissemination, or publication” of the deepfake.
Defendant cannot use a disclaimer as a defense. However, defendant could prove the plaintiff had “knowingly and voluntarily” consented to the creation, disclosure, dissemination or publication of the deepfake by signing an agreement “written in plain language that includes a general description of the sexually explicit material and the audiovisual work in which it will be incorporated,” Section 52-C 3.a. Defendant would also have to prove that such consent has not been rescinded “by delivering written notice within three business days from the date consent was given,” Section 52-C 3.b.
The new law also provides a defense to individuals who disclosed, disseminated or published a pornographic deepfake “in the course of reporting unlawful activity, exercising the person’s law enforcement duties, or hearing, trials, or other legal proceedings,” Section 52-C-4-a-I.
The law also provides a freedom of expression defense if the pornographic deepfake “is a matter of legitimate public concern, a work of political or newsworthy value or similar work, or commentary, criticism or disclosure that is otherwise protected by the constitution of [the state of New York] or the United States,” Section 52-C-4-a-ii.
Consent
The depicted individual must have been given three days to review the agreement before signing it, Section 52-C-2-b-i, unless the individual is represented by an attorney, a talent agent or a personal manager who has provided “additional written approval of the signed agreement,” Section 52-C-2-b-ii.
Therefore, if an individual is represented, the three-day delay before signing the contract may be waived, if the representative provides “additional written approval.” The law does not specify, however, the terms of this additional approval.
About the efficiency of this right of action
A court may award injunctive relief, punitive damages, compensatory damages, as well as “reasonable court costs” and attorney’s fees, Section 52-C-5. However, even putting aside the issue of litigation costs, plaintiff would face some hurdles before being awarded relief under the new law.
It may be difficult, even impossible, to identify the defendant, and even if identified, the defendant may be outside the United States, and U.S. law may not be able to establish jurisdiction.
The law could be used to file a suit against publishers of websites, some of which may be in the U.S., as it provides a cause of action against the person who disclosed, disseminated or publishes the deepfake, if knowing that the individual did not consent. However, Section 52-C.10 specifies that the law cannot be construed as limiting or enlarging the protection provided to interactive computer service providers by Section 230 of the Communications Decency Act.
It is a long and arduous process to contact websites hosting such content, and even if one succeeds in having one site take down the deepfake, another one may publish it again almost immediately.
The statute of limitation is three years “after the dissemination or publication of sexually explicit material,” Section 52-C-6-a, or one year only from the date “a person discovers, or reasonably should have discovered, the dissemination or publication of such sexually explicit material,” Section 52-C-6-b, whichever the later.
Image courtesy of Flickr user Alan Levine, Public Domain.
by 
