Students at a Beverly Hills Middle School Allegedly Created and Shared Deepfake Nudes of Their Classmates

At Beverly Vista Middle School in Beverly Hills, California, students allegedly used AI to generate fake nude photos with their classmates' faces, prompting investigations by school officials and the police. The incident highlights the increasing misuse of generative AI among minors.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.3 Fraud, scams, and targeted manipulation

The students used AI to create humiliating and sexual imagery of specific classmates for targeted harassment and cyberbullying, which directly aligns with the targeted manipulation of individuals.

Additional risk subdomains

• 1.2 Exposure to toxic content: The creation and distribution of explicit sexual deepfakes of minors exposes the victims and the school community to highly toxic and harmful content.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The incident was caused by the intentional actions of middle school students who used commercially available AI tools to generate and distribute explicit images of their classmates.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The incident involves AI-generated content such as deepfakes, which the EU AI Act classifies as Limited Risk (Level 3), requiring specific transparency obligations.

AI system and alleged parties

• AI system: None named
• AI purpose: Deepfake Image Generation; Image Generation
• Behaviour type: Tool
• Alleged developer: Unknown deepfake creator
• Alleged deployer: Unnamed middle school students
• Alleged harmed parties: Unnamed middle school students

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Substantial. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Minor, indirect Minor
• Differential treatment: direct Minor, indirect Negligible
• Civil rights: direct Minor, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Minor, indirect Negligible
• Psychological: direct Minor, indirect Minor
• Epistemic: direct Minor, indirect Negligible
• Child sexual exploitation and abuse: direct Substantial, indirect Negligible

Malicious content

Reported: Yes, the report describes the creation and distribution of AI-generated fake nude photos of students.

Directly caused: Students used AI to generate nonconsensual sexually explicit deepfake images of their classmates.

Indirectly caused: These images were shared and circulated among peers within the school community.

Inferred additional harm: It is likely that these explicit images remain on personal devices or private group chats, continuing to circulate indefinitely.

Differential treatment

Reported: Yes, the report notes that fake sexual content disproportionately harms young girls, who make up 90% of deepfake victims.

Directly caused: Female classmates were specifically targeted for sexualized deepfakes, reflecting gender-based harassment.

Inferred additional harm: The incident reinforces gender-based bullying and the sexual objectification of young girls within the school environment.

Civil rights

Reported: Yes, the report describes the violation of students' privacy, safety, and personal dignity.

Directly caused: The creation of nonconsensual sexual imagery violated the victims' basic rights to privacy and personal dignity.

Inferred additional harm: The lack of legal recourse highlights systemic gaps in protecting minors' civil rights against digital exploitation.

Privacy

Reported: Yes, the report explicitly describes the violation of students' privacy through the creation of fake nude photos.

Directly caused: Students' faces (personally identifiable information) were taken from social media and superimposed onto nude bodies without consent.

Inferred additional harm: The ease of harvesting public or semi-public photos to create explicit content poses an ongoing threat to the privacy of all students.

Psychological

Reported: Yes, the report explicitly describes psychological distress, humiliation, and feeling unsafe among the victims and students.

Directly caused: Victims experienced severe humiliation, anxiety, and fear, with students reporting they felt unsafe coming to school due to the circulation of explicit photos.

Indirectly caused: Parents and the school community experienced distress and helplessness in dealing with the digital harassment.

Inferred additional harm: It is highly likely that the targeted students suffered long-term trauma, social anxiety, and potential depression requiring therapy, affecting at least 5 to 10 individuals.

Epistemic

Reported: Yes, the report mentions that deepfakes are used to spread misinformation and create highly realistic fake depictions of real people.

Directly caused: The AI technology was used to fabricate realistic nude images of students who did not actually pose for them.

Inferred additional harm: The high realism of the deepfakes makes it difficult for observers to distinguish real images from fabricated ones, eroding trust in digital media.

Child sexual exploitation and abuse

Reported: Yes, the report discusses the incident as a potential sex crime against children and references child pornography concerns.

Directly caused: The creation of sexually explicit deepfake images of middle school students (minors) constitutes the generation of child sexual abuse material (CSAM).

Inferred additional harm: The digital replication and potential online persistence of these minor-depicting explicit images pose severe ongoing child exploitation risks.

People affected

• Occurrences reported: 1
• People reportedly harmed: 5
• People reportedly exposed: 50

Potential causes

Management

• Lack of platform safeguards: AI platforms failed to prevent users from generating nonconsensual nudes.

Technology

• Accessible deepfake consumer apps: Apps allow easy face-swapping and 'undressing' of photos.
• Sophisticated AI generation: AI technology creates highly realistic fake nude images.

Data Inputs

• Target faces from social media: Publicly shared photos provide source faces for AI manipulation.

Human Factors

• Adolescent developmental judgment: Tweens and teens have poor judgment regarding the permanence of actions.
• Lack of empathy and respect: Societal misogyny and bullying drive kids to degrade classmates.

Process and Methods

• Inadequate school education: Schools lack early training on digital consent and technology use.
• Internal handling of incidents: Administrators try to handle deepfake issues internally without police.

Regulatory Environment

• Absence of federal deepfake laws: There is no federal legislation banning or regulating deepfakes.
• Legal ambiguity for minors: State laws are unclear or difficult to apply to underage perpetrators.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts of the incident at Beverly Hills Middle School, including statements from school administrators, police, and legal experts. The nature of the AI's role as a deepfake generation tool is clear, and the resulting psychological and social harms are well-documented across multiple sources.
• Ambiguities identified: The specific AI applications used by the students are not named, and the exact number of victims is not quantified.
• Alternative interpretations: None. The event is clearly characterized as a case of AI-enabled peer-to-peer cyberbullying and nonconsensual deepfake creation.

Middle school students in California used consumer AI tools to generate and distribute nonconsensual explicit deepfakes of classmates. While the incident represents a serious violation of privacy and highlights regulatory gaps regarding AI-generated explicit material depicting minors, its national security impact is minor and manageable through local law enforcement.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Middle school students

Threat characteristics

• Imminence: Long-term. The local crisis is being handled, but the widespread availability of consumer deepfake tools represents an ongoing, long-term societal concern.
• Autonomy: Human-controlled. The AI tools were directly operated by human users to perform image manipulation based on direct inputs.
• Novelty: Established threat. The use of face-swapping and deepfake technology for harassment is an established threat with prior documented instances.

Impact by dimension

• Physical security: Negligible. The incident involved digital harassment among minors with no impact on physical systems, infrastructure, or kinetic safety.
• Information security: Negligible. The incident was a localized case of cyberbullying and did not involve state-sponsored information warfare or compromise of intelligence capabilities.
• Sovereignty: Negligible. This is a local school disciplinary and law enforcement matter with no threat to state authority or core government functions.
• Economic security: Negligible. The incident involved consumer-grade AI apps used by minors, posing no threat to strategic industries or national technological advantages.
• Societal stability: Minor. The creation of nonconsensual explicit images of minors constitutes a serious violation of privacy and civil rights, but the impact is localized and managed through standard procedures.