Can the DEFIANCE Act protect victims of deepfake nudes?
The legislation would allow victims of unconsented sexualized deepfakes to pursue civil action. However, tracking down their creators remains a challenging task.
The Senate passed the DEFIANCE Act, short for Disrupt Explicit Forged Images and Non-Consensual Edits Act, with unanimous support last week, sending the bill to the House.
If signed into law, the legislation would grant civil rights for victims to sue creators and distributors of nonconsensual deepfakes depicting intimate activity or nudity.
Survivors can sue for liquidated damages for up to $150,000 or $250,000 if the pornographic deepfake is linked to actual or attempted sexual assault, stalking, or harassment.
Complaints can be filed up to 10 years after the discovery of a deepfake or the victim's 18th birthday.
The legislation, introduced by Rep. Alexandria Ocasio-Cortez (D-NY), comes amid global outcry over unconsented artificial intelligence (AI) generated nudity images on social network X.
Thousands of women, including royals such as Catherine, Princess of Wales, and Princess Kako of Japan, were digitally undressed by xAI’s chatbot Grok.
“A big win for online safety”
Brian Long, CEO and co-founder of Adaptive Security, says the DEFIANCE Act would give victims a clearer and faster path to accountability for nonconsensual intimate AI-generated imagery.
Because there’s no uniform system across platforms for removing these images, and legal options often don’t move at the speed of online sharing, victims end up chasing copies while the content continues to circulate.
“The DEFIANCE Act is designed to close that gap by making consequences more predictable for bad actors and giving victims a clearer route to relief. In this category of abuse, speed matters as much as the final outcome,” Long told Cybernews.
The legislation was endorsed by over 80 organizations, including many that are fighting sexual violence against women and girls.
This should come as no surprise, because women are disproportionately affected. In 2023, 99% of the individuals targeted by deepfake pornography were women, according to the Security Hero report.
The Center for Democracy and Technology (CDT), a non-profit, called the Senate’s passing of the bill a “big win for online safety,” stating that the legislation addresses the unique harms of AI-generated non-consensual imagery while protecting free expression.
This isn’t the first legislation aimed at combating nonconsensual deepfakes. The Take It Down Act, signed into law in May, requires online platforms to remove AI-generated intimate depictions within 48 hours.
However, its effects are yet to be seen, as the law will only take full effect on May 19th, 2026.
Enforcement is lagging behind legislation
While the DEFIANCE Act offers hope to thousands of deepfakes’ victims, there is uncertainty about how it will be enforced in practice.
Long says attribution of non-consented deepfakes to their creators is difficult because creation is fast, accounts are disposable, and distribution is often cross-border. Meanwhile, tools to make them are widely accessible.
“A creator can post through throwaway accounts, and re-uploads can spread across services in hours. By the time a trail is identified, the material has often been copied and moved repeatedly,” Long tells Cybernews.
Ben Colman, a CEO at Reality Defender, has emphasized the need for technical infrastructure to stop sexually explicit imagery from being generated and uploaded.
“A lawsuit is, by definition, reactive. It happens after the trauma has occurred, after the images have circulated, and after the damage is done,” Colman wrote on LinkedIn.
How do I protect myself from pornographic deepfakes?
xAI announced on Wednesday that its Grok will no longer allow users to generate images of real people in revealing clothing, and the image edition will be available only to paid subscribers.
However, the Guardian investigators were able to digitally undress real women and post the content on X, suggesting the guardrails may not be working properly. Nor is Grok the only tool for creating these deepfakes.
Therefore, eliminating the possibility of becoming a target of a sexualized deepfake may not be possible, yet Long says users can take some practical steps to minimize the risk:
- Reduce the raw material available publicly. Review what photos and videos are visible on social platforms, older accounts, and public profiles.
- Limit easy voice samples. Change your voicemail greeting to the default system voice.
- Be thoughtful about the context you publish. Tighten privacy settings, remove unnecessary public profile data, and consider who can download or repost your content.
- If you’re targeted, act quickly and start documenting what you see. Save links, screenshots, and timestamps, and then report them through the platform where the content appears.
Unlock more exclusive Cybernews content on YouTube.
