Purportedly AI-Generated Video Depicting a Police Officer Punching a Person Was Reportedly Shared as Footage from French Student Protests

A purportedly AI-generated video depicting a police officer punching a person against a wall was reportedly shared on Facebook and X as footage of police violence during student protests in France. Activist Nathalie Yamb used it to condemn French authorities and call for international intervention. The clip predated the protests, and its creator acknowledged generating it with AI. One X post reportedly received more than 650,000 views.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.1 Disinformation, surveillance, and influence at scale

The AI-generated video was used by an activist to spread political disinformation on social media, aiming to influence public opinion and manipulate political processes surrounding the French government.

Additional risk subdomains

• 3.1 False or misleading information: The incident involved the generation and dissemination of false synthetic media depicting police brutality that misled social media users.

Causal factors

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

The incident was caused by a human activist intentionally sharing an AI-generated deepfake post-deployment to mislead the public and criticize the government.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The system is used to generate AI-generated content such as deepfakes, which under the EU AI Act poses a moderate risk requiring transparency obligations to ensure users are informed of its synthetic nature.

AI system and alleged parties

• AI system: None named
• AI purpose: Deepfake Video Generation; Social Media Content Generation
• Behaviour type: Tool
• Alleged developer: Synthetic media generation technology developers, Synthetic image generation technology developers, Google
• Alleged deployer: Synthetic media creators, Nathalie Yamb, Information manipulation actors
• Alleged harmed parties: X (Twitter) users, Social media users, Police departments, Law enforcement agencies, Governments, Government of France, French police, Facebook users, 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 a synthetic video depicting police brutality.

Directly caused: The AI system was used to generate a realistic video of a police officer violently punching an individual, which was designed to shock viewers.

Indirectly caused: The video was shared on Facebook by an activist to condemn the French government, exposing social media users to shocking synthetic content.

Inferred additional harm: It is likely that the video was shared further across other social media platforms, exposing thousands of additional users to the deceptive and violent synthetic content.

Epistemic

Reported: Yes, the report describes the generation and spread of a fake video of police violence that was presented as authentic.

Directly caused: An AI-generated video was created to realistically depict a police officer punching a citizen, fabricating a non-existent event.

Indirectly caused: An activist shared the video on Facebook as real footage of ongoing high school protests, misleading viewers and polluting the information ecosystem.

Inferred additional harm: The incident contributes to a broader erosion of shared reality, making it harder for the public to distinguish between real and synthetic footage of public events.

People affected

• Occurrences reported: 1
• People reportedly exposed: 1000

Potential causes

Management

• Growth Prioritized Over Safety: Flock prioritized rapid market adoption over implementing strict user controls.
• Poor Departmental Oversight: Police leadership failed to establish and enforce strict database access rules.
• Ignoring Victim Complaints: Some departments dismissed early warnings from victims regarding officer stalking.

Technology

• Optional Audit Settings: Flock made critical deterrent features like case numbers optional for agencies.
• Lack of Real-Time Alerts: The system lacked automated proactive alerts for suspicious repetitive queries.
• Broad Search Capabilities: Features like vehicle fingerprinting allowed precise tracking without plates.

Data Inputs

• Vague Input Justifications: The system accepted unverified one-word inputs like 'investigation' for searches.
• Massive Tracking Database: The system records 20 billion plate scans monthly, creating dense location data.
• Unrestricted Data Sharing: Data sharing across state and agency lines expanded the scope of surveillance.

Human Factors

• Abuse of Authority: Rogue officers leveraged their positions and system access for personal stalking.
• Victim Fear of Retaliation: Victims hesitated to report because abusers possessed weapons and legal authority.
• Unwarranted Operator Trust: System designers assumed all authorized users would act ethically without checks.

Process and Methods

• Lack of Regular Auditing: Departments rarely reviewed search logs to detect unauthorized officer queries.
• No Warrant Requirement: Officers could search any plate instantly without judicial review or a warrant.
• Flawed Reporting Paths: Victims struggled to report abuse since the police were the ones investigating.

Regulatory Environment

• Lack of Federal Laws: No federal legislation exists to govern or restrict the use of ALPR systems.
• Patchwork of State Laws: Inconsistent state laws left regulatory gaps, making prosecution difficult.
• Inadequate Legal Penalties: Some states do not treat license plate reader misuse as a standalone crime.

Information quality

• Classification confidence: High
• Reason for confidence: The report clearly identifies the origin of the video, the fact that it was AI-generated, the platform where it was shared, and the political context in which it was used. The creator's own admission on LinkedIn provides definitive proof of the AI's role, leaving little ambiguity about the nature of the incident.
• Ambiguities identified: The specific AI model or platform used to generate the video is not identified.
• Alternative interpretations: None. The evidence that the video is synthetic and was used for political influence is clear and uncontested.

An activist used a highly realistic AI-generated video depicting police brutality to criticize the French government during ongoing student protests. While the video was debunked, it demonstrates the evolving capability of synthetic media to be weaponized for domestic political disinformation and social manipulation.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: France
• Alleged perpetrator: French activist

Threat characteristics

• Imminence: Near-term. The incident occurred amidst ongoing student protests, requiring active monitoring to prevent further escalation of public unrest.
• Autonomy: Human-controlled. The AI system was used as a generation tool under the direct control of a human creator to produce the video.
• Novelty: Evolved capability. Represents an advancement in the realism of AI-generated video and its use in active political disinformation campaigns.

Impact by dimension

• Physical security: Negligible. No physical systems, infrastructure, or human safety threats were reported in connection with this incident.
• Information security: Minor. Involved the dissemination of a synthetic deepfake video to manipulate public opinion, but it was quickly debunked and lacked indicators of a coordinated state-sponsored campaign.
• Sovereignty: Minor. The video was used to criticize the French government and call for international intervention, but did not compromise core government operations or decision-making.
• Economic security: Negligible. No impact on financial systems, strategic technologies, or economic stability was reported.
• Societal stability: Minor. The synthetic footage exploited existing social tensions surrounding student protests to depict police brutality, posing a minor risk to social cohesion.