Artificial intelligence (AI) is producing hyperrealistic “digital twins” of politicians, celebrities, pornographic material, and more – leaving victims of deepfake technology struggling to determine legal recourse.
Former CIA agent and cybersecurity expert Dr. Eric Cole told that poor online privacy practices and people’s willingness to post their information publicly on social media leaves them susceptible to AI deepfakes.
“The cat’s already out of the bag,” he said.
“They have our pictures, they know our kids, they know our family. They know where we live. And now, with AI, they’re able to take all that data about who we are, what we look like, what we do, and how we act, and basically be able to create a digital twin,” Cole continued.
KEEP THESE TIPS IN MIND TO AVOID BEING DUPED BY AI-GENERATED DEEPFAKES
AI-generated images, known as “deepfakes,” often involve editing videos or photos of people to make them look like someone else or use their voice to make statements they never uttered in reality. (Elyse Samuels/The Washington Post/Lane Turner/The Boston Globe/STEFANI REYNOLDS/AFP via Getty Images)
That digital twin, he claimed, is so good that it is hard to tell the difference between the artificial version and the real person the deepfake is based on.
Last month, a fraudulent audio clip circulated of Donald Trump Jr. suggesting that the U.S. should have sent military equipment to Russia instead of Ukraine.
The post was widely discussed on social media and appeared to be a clip from an episode of the podcast “Triggered with Donald Trump Jr.”
Experts in digital analysis later confirmed that the recording of Trump Jr.’s voice was created using AI, noting that the technology has become more “proficient and sophisticated.”
FactPostNews, an official account of the Democratic Party, posted the audio as if it was authentic. The account later deleted the recording. Another account, Republicans against Trump, also posted the clip.
In the last several years, numerous examples of AI deepfakes have been used to mislead viewers engaging with political content. A 2022 video showed what appeared to be Ukrainian President Volodymyr Zelenskyy surrendering to Russia – but the fake clip was poorly made and only briefly spread online.
Manipulated videos of President Donald Trump and former President Joe Biden later appeared in the run-up to the 2024 U.S. Presidential Election. Based on existing videos, these clips often altered Trump and Biden’s words or behaviors.
AI-GENERATED PORN, INCLUDING CELEBRITY FAKE NUDES, PERSIST ON ETSY AS DEEPFAKE LAWS ‘LAG BEHIND’
A woman in Washington, D.C., views a manipulated video on January 24, 2019, that changes what is said by President Donald Trump and former president Barack Obama, illustrating how deepfake technology has evolved. (Rob Lever /AFP via Getty Images)
AI-generated images, known as “deepfakes,” often involve editing videos or photos of people to make them look like someone else by using AI. Deepfakes hit the public’s radar in 2017 after a Reddit user posted realistic-looking pornography of celebrities to the platform, opening the floodgates to users employing AI to make images look more convincing and resulting in them being more widely shared in the following years.
Cole told that people are their “own worst enemy” regarding AI deepfakes, and limiting online exposure may be the best way to avoid becoming a victim.
However, in politics and media, where “visibility is key,” public figures become a prime target for nefarious AI use. A threat actor interested in replicating President Trump will have plenty of fodder to create a digital twin, siphoning data of the U.S. leader in different settings.
CONGRESS MUST STOP A NEW AI TOOL USED TO EXPLOIT CHILDREN
“The more video I can get on, how he walks, how he talks, how he behaves, I can feed that into the AI model and I can make deepfake that is as realistic as President Trump. And that’s where things get really, really scary,” Cole added.
In addition to taking on the personal responsibility of quartering off personal data online, Cole said legislation may be another method to curtail the improper use of AI.
Sens. Ted Cruz, R-Texas, and Amy Klobuchar, D-Minn., recently introduced the Take it Down Act, which would make it a federal crime to publish, or threaten to publish, nonconsensual intimate imagery, including “digital forgeries” crafted by artificial intelligence. The bill unanimously passed the Senate earlier in 2025, with Cruz saying in early March he believes it will be passed by the House before becoming law.
First lady Melania Trump traveled to Capitol Hill on Monday for a roundtable to rally support for the Take It Down Act. (Fox News)
The proposed legislation would require penalties of up to three years in prison for sharing nonconsensual intimate images — authentic or AI-generated — involving minors and two years in prison for those images involving adults. The bill would also require penalties of up to two and a half years in prison for threat offenses involving minors, and one and a half years in prison for threats involving adults.
The bill would also require social media companies such as Snapchat, TikTok, Instagram and similar platforms to put procedures in place to remove such content within 48 hours of notice from the victim.
HIGH SCHOOL STUDENTS, PARENTS WARNED ABOUT DEEPFAKE NUDE PHOTO THREAT
First lady Melania Trump spoke on Capitol Hill earlier this month for the first time since returning to the White House, participating in a roundtable with lawmakers and victims of revenge porn and AI-generated deepfakes.
“I am here with you today with a common goal — to protect our youth from online harm,” Melania Trump said on March 3. “The widespread presence of abusive behavior in the digital domain affects the daily lives of our children, families and communities.”
Andy LoCascio, the co-founder and architect of Eternos.Life (credited with building the first digital twin), said that while the “Take it Down” act is a “no-brainer,” it is completely unrealistic to assume it will be effective. He notes that much of the AI deepfake industry is being served from locations not subject to U.S. law, and the legislation would likely only impact a tiny fraction of offending websites.
