# AI-generated non-consensual imagery and security flaws

> Developing situation record from CLSTR: https://clstr.news/situations/ai-generated-non-consensual-imagery-and-security-flaws
> Updated: 2026-11-04T00:35:06.000Z. Sources: 8. Developments: 4.

Security research has identified vulnerabilities in AI image generators from Google and OpenAI, such as DALL·E 3 and GPT-Image-2. Using the RISE method, researchers found that approximately 13% of prohibited requests were successful in bypassing safety filters to produce explicit content. Recent versions of these tools appeared more susceptible to these bypasses than previous iterations. This technical vulnerability coincides with a rise in digital violence through non-consensual sexual deepfakes. These AI-generated images and videos manipulate existing photographs to place individuals in fabricated, sexualized situations. The Internet Watch Foundation reported a 40% increase in AI-generated images categorized as child sexual abuse in the first half of 2026 compared to the entirety of 2025. Among records where age and sex could be determined, 98% depicted girls. Law enforcement agencies, such as the Noad unit of the Civil Police in São Paulo, have linked these images to extortion and threats. In Brazil, authorities noted that perpetrators, often aged between 12 and 21, use temporary accounts to evade detection. A peer-reviewed study involving over 7,200 adults in Australia, the United Kingdom, and the United States has further highlighted the prevalence of this abuse. Led by Professor Nicola Henry of RMIT University, the research found that approximately 7% of respondents have experienced AI-generated image-based sexual abuse. While men reported higher rates of experiencing this abuse, women reported more severe negative impacts, including higher rates of fear for their safety and mental health consequences. Approximately 87% of victims reported negative impacts on their wellbeing, relationships, work, or finances. Recent developments show that highly realistic, non-consensual nude images are increasingly targeting celebrities. These deepfakes often utilize real paparazzi photographs as a base, with creators replicating specific details, such as tattoos, to increase perceived authenticity. Targets have included British actor Callum Turner, singer Harry Styles, and Noah Schnapp.

## Timeline

### 2026-11-04: Study finds widespread AI-generated sexual imagery abuse

A study of 7,200 adults in the US, UK, and Australia found that 7% have experienced AI-generated image-based sexual abuse, including non-consensual creation or threats of sharing.

4 sources. https://clstr.news/cluster/study-finds-widespread-ai-generated-sexual-imagery-abuse

### 2026-11-03: AI-generated deepfakes target celebrities and flood digital spaces

AI-generated deepfakes are targeting celebrities with realistic nude imagery, while bots now account for over 57 percent of internet traffic, complicating the ability to distinguish reality from manipulation.

2 sources. https://clstr.news/cluster/ai-generated-deepfakes-target-celebrities-and-flood-digital-spaces

### 2026-10-31: AI-generated deepfakes drive rise in digital violence

The rise of AI-generated deepfakes is fueling digital violence, with a significant increase in non-consensual sexual content and AI-generated child abuse imagery reported.

2 sources. https://clstr.news/cluster/ai-generated-deepfakes-drive-rise-in-digital-violence

### 2026-10-14: AI image generators from Google and OpenAI face security flaws

Security flaws in Google and OpenAI image generators allow users to bypass filters, contributing to a rise in non-consensual AI-generated nudity and digital extortion globally.

2 sources. https://clstr.news/cluster/ai-image-generators-from-google-and-openai-face-security-flaws

---
Cite as: AI-generated non-consensual imagery and security flaws. CLSTR, https://clstr.news/situations/ai-generated-non-consensual-imagery-and-security-flaws