Purportedly AI-Generated Fake Videos of Louvre Heist Reportedly Circulated Widely Online

After the October 19, 2025 jewel heist at the Louvre, purported deepfakes were reported to have circulated widely online, allegedly depicting the robbery despite being fabricated.. The clips were reportedly posted on Facebook, Douyin and RedNote, and they were noted to have featured morphing artifacts, disappearing objects, and partial Sora watermarks. AI Forensics and independent checks reportedly found the scenes inconsistent with the real Apollo Gallery, confirming the footage was not real.

Risk classification

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

The incident involved the creation and viral dissemination of AI-generated fake videos depicting a real-world museum heist, leading to public misinformation.

Causal factors

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

The incident was caused by a human actor intentionally using OpenAI's Sora video generator to create and distribute fake footage of a real-world heist.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The system is an AI video-generation tool used to create deepfakes or AI-generated content. Under the EU AI Act, providers of such systems must ensure that outputs are marked as AI-generated, which aligns with the 'Sora watermarks' mentioned in the report.

AI system and alleged parties

• AI system: Sora (OpenAI)
• AI purpose: Deepfake Video Generation; Social Media Content Generation
• Behaviour type: Tool
• Alleged developer: OpenAI
• Alleged deployer: Unknown social media account operators
• Alleged harmed parties: social media users, Media consumers, General public of Hong Kong, General public, Epistemic integrity

Harm severity

Severity ratings are based on this scale.

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 Minor, indirect Minor
• 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 Minor
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: Yes, the report describes the creation and spread of fabricated videos depicting a crime.

Directly caused: The AI system Sora was used to generate two fabricated videos showing masked thieves breaking display cases and stealing jewelry.

Indirectly caused: These fabricated videos were shared on social media platforms, including a Facebook reel posted to an audience of over 240,000 followers, as well as on Douyin and RedNote.

Epistemic

Reported: Yes, the report describes the widespread dissemination of AI-generated fake videos depicting a real-world robbery.

Directly caused: Fabricated videos of the Louvre heist were created using AI, presenting false visual evidence of how the crime occurred.

Indirectly caused: The fake videos circulated widely on Facebook, Douyin, and RedNote, misleading viewers about the details of the heist before being debunked by fact-checkers.

Inferred additional harm: The incident likely contributed to a broader erosion of trust in online video content, as users may struggle to distinguish real news footage from AI-generated fabrications.

People affected

• Occurrences reported: 1
• People reportedly exposed: 240000

Potential causes

Management

• Prioritizing Engagement: Platforms prioritized viral reach over immediate content verification.

Technology

• Sora Video Generation: OpenAI's Sora was used to generate highly realistic fake heist videos.
• Obscurable Watermarks: Sora's watermarks were partially obscured, making identification harder.
• AI Generation Artifacts: Morphed hands and disappearing objects revealed the videos were AI-made.

Human Factors

• Deceptive Content Sharing: A user shared fake footage to a large following to gain engagement.
• Lack of Public Verification: Audiences widely shared the video without verifying its authenticity.

Process and Methods

• Inadequate Content Moderation: Social media platforms failed to detect and filter the fake heist videos.
• Weak Watermark Detection: Platforms did not scan for or flag the obscured OpenAI watermarks.

Regulatory Environment

• Lack of AI Content Mandates: No strict regulations exist to enforce labeling of AI-generated media.

Information quality

• Classification confidence: High
• Reason for confidence: The report clearly identifies the AI system involved (OpenAI's Sora) via watermarks and forensic analysis, details the platforms where the content circulated, and provides specific visual discrepancies that proved the videos were fake. There is little ambiguity regarding the events or the AI's role.

AI-generated video depicting a real-world heist at the Louvre circulated on social media, demonstrating how advanced text-to-video tools like OpenAI's Sora can be rapidly deployed to create realistic misinformation following major news events. While the incident posed negligible threats to critical infrastructure or sovereignty, it highlights an evolving capability in epistemic harm and public trust erosion.

• Overall national security impact: Minor
• Scope: Multiple nations
• Primary target: No clear primary
• Other affected: France, China
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. The immediate hoax has been debunked, but the incident highlights an ongoing, long-term challenge of rapid generative AI misinformation deployment.
• Autonomy: Human-controlled. The AI system (Sora) acted strictly as a tool, generating video content based on direct prompt inputs from a human user.
• Novelty: Evolved capability. Represents an evolution in the visual quality and speed of generating realistic video misinformation to match real-world breaking news events.

Impact by dimension

• Physical security: Negligible. The AI was used solely to generate fake videos of a heist after the fact. No critical infrastructure or physical systems were targeted or compromised by the AI.
• Information security: Minor. AI-generated deepfakes of a real-world heist circulated on social media, demonstrating the ease of creating rapid misinformation following news events, but with no targeted intelligence or state-level information warfare.
• Sovereignty: Negligible. The incident involved a museum robbery depiction and did not impact state authority, electoral processes, or core government operations.
• Economic security: Negligible. While a real-world theft occurred, the AI-generated videos did not target strategic industries, financial systems, or national technological competitiveness.
• Societal stability: Minor. The rapid spread of realistic fake videos of a crime contributes to epistemic harm and erosion of public trust in media, but did not spark civil unrest or systematic rights violations.