Taylor Swift's supposed post falsely attributed to Trump underscores the complications in regulating misuse of artificial intelligence
Anxious about the expansion of AI tools, politicians have recently initiated or advocated for initiatives to safeguard individuals from the misuse of their identity, name, voice, and visual representation in the digital era. Experts believe that Trump's social media post serves as a prime example of why these broad legislative actions are being introduced, alongside being one of the most prominent fabricated statements in the 2024 race.
Last weekend, Trump reposted a series of images on his Truth Social platform, which included an image depicting Taylor Swift as Uncle Sam. The post included the caption, "Taylor wants you to vote for Donald Trump." In response to this alleged endorsement, Trump posted a slideshow of Swift images, accompanied by the comment, "I accept!"
A representative for Swift did not promptly respond to CNN's request for comment. Swift has not endorsed any candidate in this election.
In an interview with Fox Business on Wednesday, Trump denied creating the images when asked if he was concerned about a potential lawsuit.
"I'm unaware of anything about them, other than someone else producing them, I didn't produce them," Trump stated. "These were all created by others. AI is consistently perilous in this manner."
The ELVIS Act comes into play
Tennessee is one of the most recent states to implement a law aimed at shielding individuals from unauthorized use of content that mimics their visual representation or likeness.
The Ensuring Likeness Voice, and Image Security Act (ELVIS Act), which went into effect last month, broadens Tennessee's existing right-of-publicity law to specifically protect artists, including a person's voice, and declares unlawful the use of content "in any medium."
The legislation could potentially enable Swift – who began her career in Nashville, where she has a part-time residence – to sue.
Tennessee Senate Majority Leader Jack Johnson, a Republican who sponsored the bill, has mentioned that the misuse of AI-generated content and its impact on artists were among the reasons for updating the law.
“The rapid advancement of AI is exhilarating in many ways, but it also presents new challenges – especially for singers, songwriters, and other music professionals,” Johnson said in a press release when Tennessee Gov. Bill Lee signed the ELVIS Act into law. CNN has reached out to the governor's office for comment.
Joseph Fishman, a law professor at Vanderbilt University whose research has centered on intellectual property and entertainment law, has pointed out that one of the issues with the law is that it's so extensive that "it covers virtually any unauthorized use of a person's likeness or voice that the distributor of the image or video or sound knows wasn't authorized."
Since 2019, several states have enacted legislation related to the use of fake content. In the 2024 session, at least 40 states have pending legislation, according to the National Conference of State Legislatures. Although the laws do not exclusively apply to content generated by AI, many laws aim to address sexually explicit content and some focus on content intended to mislead voters, the group said.
When it comes to political campaigns, more than a dozen states have enacted legislation to oversee the use of "deepfakes" – realistic fake video, audio, and other content created with AI. Depending on the state, violators could face imprisonment and heavy fines; candidates could be required to surrender their office or nomination, CNN previously reported.
Many laws, but no federal statute
While there seems to be a surge in AI-related legislation, the unauthorized use of digital replicas could potentially be punishable under existing laws, stated Corynne McSherry, the legal director at the nonprofit Electronic Frontier Foundation, who specializes in intellectual property, open access, and free speech issues.
“If you’re concerned about the unauthorized use of an image of you, or your face, and you think it's defamatory or implies a false endorsement, you likely already have rights under defamation law, potentially even under trademark law, as we have many longstanding doctrines to address such situations,” McSherry said.
At the federal level, Congress has yet to pass a comprehensive framework to manage AI, including AI-generated replicas. However, the Federal Communications Commission has sought fines following an AI-generated robocall imitating President Joe Biden's voice that urged voters not to participate in New Hampshire's primary election. The robocalls utilized call-spoofing technology that violated federal caller-ID laws, the FCC stated.
The carrier that transmitted the robocalls, Lingo Telecom, agreed on Wednesday to pay a $1 million fine. Steven Kramer, the political consultant behind the call, faces a $6 million fine.
Last month, the US Copyright Office published a report urging lawmakers to pass a federal law to address unauthorized digital replicas.
“It has become apparent that the distribution of unauthorized digital replicas poses a significant threat not only in the entertainment and political arenas but also for private citizens,” Shira Perlmutter, register of copyrights and director of the US Copyright Office, said in a statement. “We believe there is an urgent need for effective nationwide protection against the damages that can be inflicted on reputations and livelihoods."
For Darrel Mottley, a patent attorney and faculty director of the intellectual property and entrepreneurship clinic at Suffolk University, it's crucial to remember that regulation should focus on how AI is utilized – not the technology itself.
“We don’t regulate the technology, per se, but rather the human behavior of utilizing the technology inappropriately. That's what the regulations should be doing,” Mottley said.
Could Taylor Swift sue Trump?
Legal experts believe that Swift could file a lawsuit under the ELVIS Act due to her connections to the Southern state, but the outcome remains uncertain.
"The ELVIS Act could potentially be one of the regulations that Trump could be held accountable for, according to Fishman's statement.
In a possible legal battle against Trump, his legal team might contend that the post was an expression of satire or parody, which is shielded under the First Amendment. Fishman and McSherry shared this perspective.
The ELVIS Act contains a provision for uses safeguarded by the First Amendment, such as criticism, satire, and parody. Fishman pointed out, however, that the limitations of this provision weren't clarified in the legislation, leaving room for uncertainty about how courts will interpret it.
Fishman stated, "There's a lot of ambiguity concerning how this scenario would unfold if there was a lawsuit, but should this exemption not apply, sharing these images appears to fall under this Tennessee law."
With some images being produced by AI, the situation became more intriguing, Fishman mentioned. However, Trump could still be subject to the same legal repercussions "if he edited a photo using Photoshop (or) if he was an exceptionally skilled artist and sketched something. The issue remains the same."
In McSherry's opinion, Swift could avoid litigation and choose to publicly denounce the content instead.
"Considering Taylor Swift's considerable celebrity influence, I believe she could achieve far more impact, honestly, by just using her own platform to distance herself from it. This would basically achieve the same result as a lawsuit," McSherry suggested."
Regardless of the widespread use of AI-created content, legal professionals agreed that individuals should not be spared from accountability when their voice, appearance, or image is utilized to suggest false endorsement, even if they're not famous.
The ELVIS Act in Tennessee could potentially allow Taylor Swift to sue Trump if she deems his use of her image as a form of unauthorized use of her visual representation or likeness. According to Joseph Fishman, the law is so extensive that it covers unauthorized uses that the distributor knows weren't authorized.
In an interview, Trump stated that AI is consistently perilous in this manner, suggesting that he was unaware of the AI-generated images used in his posts. However, the use of AI in creating such content does not exempt individuals from accountability when their voice, appearance, or image is used to suggest false endorsement.
