In an era where digital privacy is increasingly under threat, the recent deepfake incident involving the globally renowned artist, Taylor Swift, has catapulted the issue of nonconsensual pornography to center stage. This high-profile case, that involved fabricated explicit images of Swift being circulated on social media without her consent, has sparked a wave of public indignation and a clamor for legal reform. Beyond the headlines, this incident exposes a grim reality faced by countless individuals who fall victim to what is commonly known as "revenge porn." Despite the issue’s growing prevalence, the legal system doesn’t always offer many victims a clearcut path to recourse.
The surge in the use of mainstream artificial intelligence (AI) software has raised heightened concerns, including for social media platforms, given AI’s capacity to generate remarkably authentic and potentially harmful images. The challenge lies in developing effective content moderation policies without infringing upon the broader ethos of free expression (such as political satire and other free speech that the First Amendment protects). Despite considerable efforts by social media companies and state law, significant gaps remain in the legal protections offered to victims of nonconsensual pornography. The emergence of deepfake technology presents new legal challenges, as exemplified by the Taylor Swift case, where traditional definitions of nonconsensual pornography may not suffice.
The Swift Incident: A Turning Point
The incident involving Taylor Swift, a globally recognized celebrity, marks a significant turning point in public sentiment over nonconsensual pornography. Swift, known for her immense fan following and influential voice, became an unwitting victim when deepfake technology was maliciously used to create and disseminate explicit images of her without her consent. These images, though fabricated, rapidly spread across social media platforms, attracted more than 45 million views, 24,000 reposts, and hundreds of thousands of likes and bookmarks for nearly 17 hours before being shut down by X (formerly Twitter), igniting a firestorm of media coverage and public outcry. Although X was able to block “Taylor Swift” as a search term, some of the false images of Swift continued to circulate the social media platform, because individuals were able to bypass the search block by manipulating search terms, such as adding words between the pop star's first and last name.
What is Deepfake Technology?
Deepfake Technology refers to a form of artificial intelligence utilized for fabricating persuasive images, audio, and video deceptions. This term encompasses both the underlying technology and the fabricated content, derived from a fusion of deep learning and deception. Deepfakes typically involve altering pre-existing source material by substituting one individual with another. Moreover, they generate entirely new content portraying individuals engaging in actions or uttering statements they never actually performed or said. The primary peril associated with deepfakes lies in their capacity to disseminate misinformation that seemingly originates from reliable outlets.
State Initiatives to Combat Deepfake Technology
Laws addressing non-consensual pornography directly apply in instances where deepfakes are employed to generate explicit material without the subject's consent. Yet, the implementation of these laws can be hindered by jurisdictional complexities and the anonymous nature of online content dissemination. Several states have enacted some form of legislation targeting this conduct. However, these laws are far from uniform, with significant variations in definitions, protected activities, and penalties. For instance, some states focus only on images obtained unlawfully or with an intent to harm, while others include broader protections. The penalties range from misdemeanors to felonies, reflecting differing perceptions of the severity of the crime. For example, some states, such as New York, address only the dissemination of such material, rather than the creation of it.
Laws requiring consent for use of an individual’s likeness are complicated by First Amendment freedom of speech protections. Therefore, in most jurisdictions, consent requirements depend on factors such as whether the use is commercial or whether the individual depicted is a celebrity. However, enforcing even these limited consent requirements is challenging with deepfakes as they are normally created anonymously and distributed widely.
In addition to a scattered legal landscape, some technology offerings may increase the potential harms from deepfakes. Internet platforms apply various levels of content moderation, so some platforms facilitate photo and video sharing, where deepfakes can be used for defamation, blackmail, and other malicious purposes. Content moderation can be especially difficult on live steams. That lack of moderation is exacerbated by the anonymity and ease of creating and disseminating deepfakes, making it exceedingly difficult to identify the creators and hold them accountable. These factors converge to create an environment ripe for the misuse of this technology.
Social Media Platform Immunity
Due to the protections granted under Section 230 of the Communications Decency Act, holding social media platforms legally accountable for their users’ distribution of deepfakes presents challenges. Section 230 of the Communications Decency Act has been a cornerstone of internet law in the United States since its enactment in 1996. This legislation offers fairly wide immunity to online platforms from liability for content posted by their users, fostering a free and open internet. However, in the context of nonconsensual pornography, especially with the emergence of deepfake technology, the provisions of Section 230 necessitate a nuanced analysis to balance the rights and responsibilities of online platforms.
As it stands, Section 230 provides broad immunity to online service providers regarding third-party content. This immunity has been crucial in allowing platforms such as social media sites, forums, and comment sections to flourish without the constant threat of litigation. However, this protection also means that platforms have limited legal incentive to address the spread of nonconsensual pornography proactively (although many have chosen to do so in the interest of creating a safe platform). Last May, the Judiciary Committee advanced a package of five bills related to online child sexual abuse material (CSAM). One bill, Earn It Act, would roll back Section 230 protections when platforms facilitate content that violates civil and state criminal laws on child sexual exploitation. Another bill, the Stop CSAM Act, would create a new cause of action for victims and their families to sue over such material. Neither of these bills have made it to the Senate floor for a vote.
Additionally, the Supreme Court is currently deliberating a crucial issue regarding the extent of state authority over social media platforms.1 Specifically, the focus is on whether states such as Florida and Texas have the jurisdiction to mandate these platforms to carry content that they deem hateful or objectionable. The state laws explicitly allow users to sue tech platforms for alleged censorship that Section 230 shields them from. Although deepfake non-consensual pornography was not directly addressed in oral arguments, altering the scope of Section 230 could potentially broaden the scenarios in which social media platforms could face litigation for that content.
Proposed Amendments
On January 30, 2024, the Senate introduced a bipartisan bill, the Disrupt Explicit Images and Non-Consensual Edits Act of 2024 (“DEFIANCE”), intended to hold accountable those who are responsible for the proliferation of nonconsensual, sexually explicit deepfake images and videos. The civil remedy applies to digital forgeries that depict the victim in the nude or engaged in sexually explicit conduct of sexual scenarios. This remedy is enforceable against individuals who produced or possess the forgery with intent to distribute it; or who produced, distributed, or received the forgery, if the individual knew or recklessly disregarded that the victim did not consent to the conduct.
The one pager of the legislation specifically mentions the Swift incident: “Sexually-explicit deepfake content is often used to exploit and harass women—particularly public figures, politicians, and celebrities. For example, in January 2024, fake, sexually explicit images of Taylor Swift that were generated by artificial intelligence swept across social media platforms. Although the imagery may be fake, the harm to the victims from the distribution of sexually explicit deepfakes is very real. Victims have lost their jobs and may suffer ongoing depression or anxiety.”
This incident is not just another case of a celebrity targeted by digital abuse; this incident may represent a watershed moment in highlighting the pernicious and widespread issue of deepfake non-consensual pornography. The Swift case has succeeded in drawing attention to the severe implications of such acts, not just for public figures, but for individuals across all walks of life. It underscores the ease with which technology can be abused to invade privacy and inflict harm, raising questions about the effectiveness of existing legal protections against sophisticated forms of digital abuse such as nonconsensual pornography, including deepfakes, while protecting the virtues of free expression and technological innovation.
The below chart outlines the states with legislation that specifically target deepfake content including what conduct is prohibited, the penalties for violations, and explicit references to Section 230 protections for social media platforms.
Footnotes
1 NetChoice, LLC v. Paxton and Moody v. NetChoice, LLC.
