Second Consecutive Year of Alleged AI-Generated Images Depicting Katy Perry at Met Gala Circulating Online

Images widely reported to be AI-generated appeared to depict Katy Perry attending the Met Gala and circulated broadly online, which appeared to prompt confusion among some fans. Perry later clarified via Instagram that she had not attended the event. While no malicious actor or harm has been confirmed, this marks the second consecutive year such images have gone viral

Risk classification

• Primary risk domain: 3 Misinformation
• Primary risk subdomain: 3.1 False or misleading information

The AI system was used to generate highly realistic but false images of Katy Perry at the Met Gala, misleading her mother and fans into believing she attended the event.

Causal factors

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

The incident was caused by human users intentionally generating and distributing deepfake images of Katy Perry using deployed AI image generation tools.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The report describes AI-generated content (deepfakes) which pose a moderate risk and require transparency obligations to ensure users are informed.

AI system and alleged parties

• AI system: None named
• AI purpose: Deepfake Image Generation; Image Generation
• Behaviour type: Tool
• Alleged developer: Unknown deepfake technology developers
• Alleged deployer: unknown
• Alleged harmed parties: Truth, Katy Perry, General public, Fans of Katy Perry

Harm severity

Highest direct severity in any category: Minor. 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 Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Epistemic

Reported: The report explicitly describes epistemic harm where AI-generated images misled fans and the singer's mother into believing she attended the Met Gala.

Directly caused: The viral AI-generated images directly misled Katy Perry's mother and numerous fans on social media, causing temporary confusion about her attendance at the 2024 and 2025 Met Gala.

Indirectly caused: N/A

Inferred additional harm: It is likely that thousands of social media users who viewed the viral images were temporarily misled before the singer clarified her absence.

People affected

• Occurrences reported: 2
• People reportedly harmed: 2
• People reportedly exposed: 2

Potential causes

Management

• Prioritizing Engagement over Safety: Platforms prioritized user engagement over implementing deepfake detection.

Technology

• Highly Realistic Image Generation: AI models generated highly realistic images matching the Met Gala themes.
• Lack of AI Watermarking: Generated images lacked embedded watermarks to identify them as AI-created.

Data Inputs

• Abundant Public Training Data: Plentiful public images of the celebrity allowed accurate model training.
• Met Gala Theme Data Availability: Contextual data about the event theme was used to make images plausible.

Human Factors

• User Susceptibility to Deception: Social media users and family members easily believed realistic deepfakes.
• High Trust in Visual Media: Viewers assumed photos were real without verifying the celebrity's location.

Process and Methods

• Ineffective Platform Filtering: Social media platforms failed to flag or label the viral AI images.
• Rapid Viral Distribution: Algorithms promoted highly engaging deepfakes quickly across networks.

Regulatory Environment

• Lack of Deepfake Regulations: Absence of strict laws governing the creation of non-consensual deepfakes.

Information quality

• Classification confidence: High
• Reason for confidence: The report clearly details the creation and viral spread of the AI-generated images, the reactions of Katy Perry and her mother, and the context of her actual whereabouts. There is no ambiguity about the nature of the images or the confusion caused.
• Ambiguities identified: The specific AI tools used to generate the images are not identified.

Viral AI-generated images of singer Katy Perry at the Met Gala circulated online, temporarily misleading fans and family. While demonstrating the realistic capabilities of generative AI, the incident was benign, non-malicious, and carries negligible national security implications.

• Overall national security impact: Negligible
• Response level: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents a long-term strategic and cultural challenge regarding synthetic media rather than an active national security crisis.
• Autonomy: Human-controlled. Humans actively initiated, generated, and distributed the images using AI image generation tools.
• Novelty: Established threat. AI-generated celebrity deepfakes are an established phenomenon with multiple prior occurrences.

Impact by dimension

• Physical security: Negligible. No physical threats, kinetic attacks, or critical infrastructure manipulation occurred.
• Information security: Negligible. The incident involves benign viral pop-culture images without malicious intent, state-actor involvement, or systematic information warfare.
• Sovereignty: Negligible. No government functions, electoral systems, or sovereign decision-making processes were impacted.
• Economic security: Negligible. No threat to financial systems, critical supply chains, or strategic technological competitive advantages.
• Societal stability: Negligible. While the deepfakes temporarily misled public figures and fans, they caused no societal instability, mass surveillance, or human rights violations.