New Research from Stanford Internet Observatory Documents Nudification Tool Usage Targeting High School Students

Researchers at the Stanford Internet Observatory published a study on October 16, 2026, documenting the use of AI-powered nudification applications to create non-consensual sexualized images of high school students, based on analysis of 847 reports from schools, law enforcement, and platform trust-and-safety teams across the United States between January 2025 and September 2026. The study found that in 68 percent of cases, the manipulated images were shared via Instagram, Snapchat, or Discord direct messages to the victim or to peer groups, often accompanied by blackmail demands or used as a tool of harassment. The research identified nudification as a distinct vector of technology-facilitated abuse, noting that existing content-moderation systems struggle to detect synthetic images that contain no actual photographic CSAM.

Detection of the sharing event, rather than reliance on image-scanning alone, addresses the operational gap the Stanford research identifies. Guardii's world-leading platform monitors direct-message exchanges for the combination of synthetic intimate imagery and the coercion, threat, or humiliation language that the study found in the majority of nudification cases involving minors. By analyzing conversation patterns across Instagram, Snapchat, Discord, and other channels where the Observatory documented distribution, the system flags high-risk exchanges for intervention before a manipulated image reaches its target or circulates within a peer group, and surfaces the incident to parents, schools, or platform safety teams for immediate response. The approach detects harm that image-only filters miss, directly targeting the 68 percent of cases the researchers traced to direct-message platforms.