Two men in the U.S. have been federally charged with using artificial intelligence to generate sexually explicit photos and videos of women without consent under a new law, as global concern grows over regulating misuse of A.I.
Arturo Hernandez, 20, and Cornelius Shannon, 51, were arrested in unconnected cases on Wednesday and each faces up to two years in prison. The men are some of the earliest defendants charged under the 2025 Take It Down Act, which criminalizes the publication and spread of non-consensual A.I.-generated deepfakes.
Governments around the world are scrambling to develop safeguards as the use of generative A.I. grows more common, including to spread misinformation and non-consensual images. Lawmakers and victim-advocacy groups have for years sounded the alarm on the rise of “revenge porn,” and expressed concern that A.I. is making such exploitation easier and more widespread. Almost every U.S. state now has laws regulating deepfakes or has introduced legislation to do so.
“The defendants used cutting-edge digital technology to create images that degraded and violated victims across the United States,” U.S. Attorney for the Eastern District of New York Joseph Nocella said in a Wednesday statement. “This case makes clear that posting deepfake pornography is not a victimless crime.”
Here’s what to know about the law and who has been charged so far.
What is the Take It Down Act?
The Take It Down Act, is the first major federal law in the U.S. to address A.I.-related harms. Signed into law by President Donald Trump last May, the act makes it a federal crime to knowingly publish or threaten to publish non-consensual intimate imagery, including both real images and A.I.-generated deepfakes, and it mandates platforms remove the material within 48 hours of being notified.
The bill was introduced by Sen. Ted Cruz (R, Texas) and Sen. Amy Klobuchar (D, Minn.) and received broad bipartisan support, including the backing of First Lady Melania Trump. While some civil society groups raised concerns that the bill might lead to censorship, it has largely been hailed as an important first step toward providing legal recourse against the spread of both A.I.-generated and real non-consensual images.
“We must provide victims of online abuse with the legal protections they need when intimate images are shared without their consent, especially now that deepfakes are creating horrifying new opportunities for abuse,” Klobuchar said last February. “These images can ruin lives and reputations, but now that our bipartisan legislation is becoming law, victims will be able to have this material removed from social media platforms and law enforcement can hold perpetrators accountable.”
“This is also a landmark move towards establishing common-sense rules of the road around social media and A.I.,” she said last May.
How big of a problem is A.I. “revenge porn”?
The measure was spurred by the activism of a Texas high schooler, Elliston Berry. In October 2023, Berry, then-14, discovered that a classmate had used A.I. software to make and share a deepfake nude image of her. Berry appealed to Snapchat to remove the images but the social media platform refused for almost a year. Finding limited avenues to seek accountability, Berry and her mother visited Cruz’s office to ask for help.
Around the same time, Francesca Mani, then a 15-year-old from New Jersey, also learned that a classmate had fabricated sexually explicit images of her and other female classmates using A.I. software. She joined Berry in advocating for legislation that would protect them and penalize the perpetrators.
Berry and Mani aren’t the only victims of deepfake nudes.
In March, three teenagers in Tennessee filed a lawsuit against Elon Musk’s xAI, alleging that the company’s A.I. tool Grok had created fake sexually explicit images out of real images of them without their knowledge.
Last August, xAI released Grok Imagine, which included a “spicy mode” setting that allowed paying users to prompt the chatbot to generate fake sexual images, including the “undressing” of real people. Grok produced around three million sexualized images, including 23,000 of children, within weeks of the release, according to an analysis by the Center for Countering Digital Hate. xAI has since restricted “spicy mode” to filter out sexually explicit requests.
Also in March, two teenage boys in Lancaster, Pa., received probation after they created hundreds of fake nude photos of dozens of their female classmates, most of whom were under the age of 18. The boys had prompted A.I. software to morph real images of the girls taken from school photos, yearbooks, and social media into sexually explicit images, according to authorities.
A number of female celebrities have also spoken out about the distribution of deepfake images of them. Pop star Taylor Swift filed trademark applications in April in an apparent effort to protect her voice and likeness from deepfakes after countless non-consensual A.I.-generated images of her circulated the internet, including sexually explicit images and fake political endorsements.
What are these cases about?
Court filings alleged that Shannon, who is from New Jersey, published at least 360 albums containing A.I.-generated pornography of around 90 different women since last May. These materials featured female politicians and celebrities and were shared to an internet platform where they received millions of views, the complaint said.
Hernandez, from Texas, is alleged to have published at least 113 albums of deepfake pornography featuring at least 50 women, including both celebrities and women who are not public figures, since last May. The materials appear to be of real non-explicit images of women that were morphed into sexually explicit images, according to the complaint. These were published on a website where they received nearly a million views. Some of the materials were deepfakes of recent high school graduates, the complaint said.
Both men were charged in federal court in Brooklyn on Wednesday.
“This predatory conduct represents a disturbing abuse of technology that inflicts emotional harm on victims, violating their privacy, dignity, and security,” FBI Assistant Director in Charge James C. Barnacle Jr. said in a statement. “The use of this emerging technology to victimize individuals is not innovative—it is criminal and will be pursued with the full force of the law.”
Have there been previous cases under the law?
Last month, 37-year-old James Strahler II became the first person to be convicted under the newly-enacted law. Strahler, from Ohio, pleaded guilty to using A.I. to create non-consensual sexually explicit images and videos of both adults and minors.
Strahler had posted more than 700 A.I.-generated pornographic images of both real and animated people to a website dedicated to child sexual abuse, and he had used A.I. software to create sexually explicit images of minor boys from his community and several adult women, according to the Department of Justice. He is awaiting sentencing.
