Introduction
Deepfakes are audio or visual impersonations generated by Artificial Intelligence (AI) that replicate a person’s voice, face, or likeness. While they can be used harmlessly for creative and entertainment purposes, deepfakes also come with serious risks, including the spread of misinformation, identity theft, and the exploitation of people through non-consensual intimate content.
As generative AI has grown in popularity, deepfake creation tools have become more accessible. This has resulted in an increase in harmful uses of AI, including deepfake pornography, which makes up 98 percent of all deepfake videos online. According to a 2023 study, nearly half of the surveyed men reported having come across deepfake pornography at least once in their lives. The study also revealed that 99 percent of individuals targeted by deepfake pornography are women.
The Take It Down Act
Lawmakers have introduced a number of bills to address the misuse of deepfakes. In January 2025, Senators Ted Cruz (R-TX) and Amy Klobuchar (D-MN) reintroduced the Take It Down Act, or TIDA (S. 146), after it was first introduced in June of 2024. The bill passed in the Senate unanimously in February 2025 and was later approved by the House before being signed into law by President Trump in May 2025.
Under the law, individuals have the right to request the removal of intimate deepfake content or non-consensual sexually explicit content involving minors. In addition, online platforms must remove such content if they are notified of its existence. The law intends to give victims legal options while placing responsibility on tech companies to act fast in removing harmful content.
However, TIDA was not the first attempt at federal regulation of deepfakes. During the Biden administration, lawmakers introduced the Defiance Act, which aimed to address similar concerns around image-based sexual abuse and non-consensual deepfake distribution. While the bill was passed by the Senate, it fell short in the House, preventing it from being signed into law. Its failure highlighted the political and practical difficulties of regulating deepfakes; while there is an urgent need to protect victims of digital exploitation from the psychological and reputational harm caused by deepfakes, the right to free speech and expression must also be preserved. The difficulty in striking a balance between these two things forms the central conflict surrounding AI regulation. Clearly, the ever-evolving nature of AI means that regulatory measures for the technology must be handled with care.
Pros of TIDA
Supporters of deepfake regulation have argued that laws like the Take It Down Act are necessary steps to protect vulnerable individuals, especially children and young adults, from exploitation and psychological harm. Victims are prone to suffer long-term mental health consequences, including anxiety, depression, and even suicidal thoughts. As one grieving parent stated, “Victims of these crimes often suffer from mental health trauma from the fear that these images will stay online forever,” highlighting the critical need for protection.
The Act equips schools and educators with the means to respond to student-related situations. During the 2023-2024 academic year, 71 percent of teachers reported instances of students who were caught spreading deepfakes, highlighting the expanding issues within educational settings. The law aims to reduce victimization and hold offenders accountable by mandating that online platforms swiftly remove harmful content. Furthermore, similar state-level laws, like California’s 2024 bill that criminalized the creation of deepfake images that depicted intimate body parts, demonstrate a growing consensus on the need for legal frameworks to combat these issues.
Cons of TIDA
Despite its implementation, TIDA faces challenges similar to those of the Defiance Act. Critics of TIDA have raised concerns about unforeseen repercussions, especially concerning free speech and censorship. Some argue that this law will violate First Amendment rights by permitting excessive enforcement mechanisms that could be unconstitutional and might lead to the suppression of users’ freedom of self-expression.
There are also practical challenges in identifying and verifying deepfakes. According to some experts, detection techniques often fall behind the rapid advancements of deepfake creation technologies, making enforcement difficult. Additionally, there have been concerns that the takedown process could be abused, leading to bogus reports that silence legitimate voices or suppress protected content.
Free speech advocates and digital rights groups have also argued that if the law is not applied carefully, the bill will risk being too broad. For example, many critics believe that the bill’s imprecise terminology may unintentionally ban LGBTQ+ content or lawful adult materials. These concerns highlight how challenging it is to create laws that are meant to protect victims without restricting freedom of expression or unintentionally harming marginalized communities.
Conclusion
The debate surrounding deepfake regulation, demonstrated in the discussion around the Take It Down Act, reflects the challenging balance between protecting individuals from harm and upholding civil liberties. Although victims of non-consensual deepfake abuse benefit greatly from the law, issues with censorship, free speech, and enforcement remain unresolved.
Moving forward, policymakers must continue improving on these regulations to ensure they are effective, fair, and compliant with the Constitution. As technology evolves, so must our legal and ethical framework progress, ensuring that innovation does not grow at the cost of American’s safety.