White: AI deepfakes fuel violence against women
An opinion column in The Rocky Mountain Collegian argues that AI generated sexual deepfakes are becoming a new form of technology enabled violence against women, allowing realistic intimate images and videos to be created without the victim ever having produced sexual content. The author emphasizes that once such material is distributed online, victims may have little control over where it spreads or how long it remains accessible.
The article highlights the strongly gendered nature of the problem, citing data claiming that 96 percent of deepfakes shared online are nonconsensual sexual deepfakes and 99 percent of those depict women. Rather than primarily being used for political misinformation, the author argues that deepfake technology is increasingly being used to humiliate, sexualize and damage the reputations of ordinary people, including young women.
A prominent example involves Grok on X. Earlier in 2026, users began asking the AI system to manipulate photographs of fully clothed women to depict them wearing bikinis. According to figures cited from The Guardian, the trend escalated rapidly, with Grok receiving as many as 6,000 requests for bikini videos per hour by January 8.
Some requests went significantly further than sexualization. Users reportedly asked the AI to add blood, injuries, scars, bruises and other violent elements to images of women. The author argues that this combination of sexualization and simulated violence demonstrates how generative AI can be used not merely for fabricated pornography but also for targeted intimidation and degradation.
The column raises particular concerns about children and teenagers. Because a convincing sexual deepfake can potentially be generated from an ordinary photograph within seconds, minors can become targets without ever having created or shared intimate material themselves. The author argues that sexually explicit synthetic material depicting children should be treated as child sexual abuse material and subject to appropriate legal consequences.
The article also discusses gaps in US legislation. It says that while various states have enacted laws addressing the distribution of deepfake sexual imagery, fewer laws specifically address its creation. It points to California requirements concerning disclosure of AI generated content and Minnesota legislation targeting nudification technology as examples of governments beginning to respond.
The author argues that responsibility should not rest primarily with victims. Instead, greater pressure should be placed on technology companies and online platforms to prevent their systems from facilitating abusive deepfakes, while legislation needs to evolve alongside increasingly capable generative AI tools.
Overall, the column presents nonconsensual sexual deepfakes as a problem extending beyond misinformation or reputational damage. It frames them as a serious issue involving privacy, consent, sexual harassment, misogyny and technology enabled abuse, particularly affecting women and increasingly younger victims.
