White House Reportedly Shares Purportedly AI-Altered Arrest Photo Depicting Minnesota Protester Nekima Levy Armstrong as Crying

The White House reportedly posted a purportedly AI-altered image on X showing Minnesota protester and attorney Nekima Levy Armstrong appearing to cry during her arrest. An earlier image reportedly shared by Homeland Security Secretary Kristi Noem showed her calm. Reported analysis by third parties using AI detection tools found signs of purported facial manipulation, which could be replicated using generative AI systems.

Risk classification

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

The White House used AI-manipulated media on official social media channels to conduct targeted political influence and shape public opinion regarding a high-profile arrest.

Additional risk subdomains

• 1.1 Unfair discrimination and misrepresentation: The protester's skin was darkened and her facial expression altered to depict her crying, representing her unfairly and potentially prejudicing her legal case.

Causal factors

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

The incident was caused by human actors in the White House intentionally using generative AI tools post-deployment to alter and distribute a misleading image for political messaging.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The incident involves 'AI-generated content such as deepfakes' which falls under Limited Risk due to transparency obligations to notify users when AI-generated content is used.

AI system and alleged parties

• AI system: Gemini, Grok, Resemble.AI (Google, Resemble.AI, xAI)
• AI purpose: Deepfake Image Generation; Image Retouching
• Behaviour type: Tool
• Alleged developer: Unknown image generator developers, Unknown deepfake technology developers
• Alleged deployer: White House Communications Team, White House, Executive Office of the President
• Alleged harmed parties: Nekima Levy Armstrong, 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 Negligible
• Differential treatment: direct Minor, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Minor, indirect Minor
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: The report explicitly describes the creation and spread of a manipulated image of an arrestee by the White House on X.

Directly caused: The White House posted a digitally altered image of Nekima [PERSON_001] appearing to sob with darkened skin on X.

Inferred additional harm: The post was likely shared and viewed by millions of social media users, spreading the misleading and malicious depiction widely.

Differential treatment

Reported: The report explicitly describes the darkening of the protester's skin in the altered image, which suggests racialized misrepresentation.

Directly caused: The White House posted an altered image where the black protester's skin appeared to have been darkened and her face manipulated to show her sobbing.

Inferred additional harm: This manipulation could reinforce racial stereotypes and lead to biased public perception and unequal treatment of the protester.

Democracy

Reported: The report explicitly describes official government accounts distributing AI-generated and manipulated content to influence public perception.

Directly caused: The White House and Homeland Security Secretary posted and shared manipulated imagery of an arrest on official social media channels.

Indirectly caused: The use of 'memes' and altered media by the executive branch to mock or discredit protesters undermines trust in public institutions and the rule of law.

Inferred additional harm: Widespread adoption of synthetic propaganda by state actors erodes democratic accountability and the integrity of official communications.

Epistemic

Reported: The report explicitly describes the creation and distribution of a fake, digitally altered image representing a real-world event.

Directly caused: The White House posted a version of the arrest image where the subject appears to be sobbing and her skin is darkened, which was proven false by AI detection and original photos.

Inferred additional harm: The incident contributes to the erosion of shared truth, making it harder for the public to distinguish between real and synthetic evidence of government actions.

People affected

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

Potential causes

Management

• Condoning Synthetic Media Use: Leadership defended the post, stating 'the memes will continue'.
• Lack of AI Communication Policy: No management oversight to restrict the use of AI-manipulated images.

Technology

• Generative AI Image Modification: Generative AI tools enabled realistic alteration of the original arrest photo.
• AI Face and Skin Manipulation: AI algorithms modified facial expressions and darkened skin tones easily.

Data Inputs

• Real Arrest Photo as Input: Original authentic photo was fed into AI tools to generate the fake version.

Human Factors

• Intentional Distribution of Memes: Officials chose to distribute AI-altered images for political communication.
• Enthusiastic Sharing of AI Content: Key figures actively promoted synthetic and manipulated media.

Process and Methods

• Lack of Media Verification: No formal process was used to verify the authenticity of official posts.

Information quality

• Classification confidence: High
• Reason for confidence: The report clearly details the incident, the specific AI tools used to detect and replicate the manipulation, and the political and legal implications. There is little ambiguity regarding the facts of the image alteration.
• Ambiguities identified: The exact AI tool used by the White House staff to perform the original manipulation is not explicitly named, though Gemini and Grok were shown to be capable of it.

The White House's distribution of an AI-manipulated image of a protester represents a minor national security concern. While it does not threaten physical infrastructure, it undermines information integrity from official channels and introduces legal complications that could hinder a federal prosecution, highlighting the risks of state-level adoption of synthetic media for political messaging.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: The White House

Threat characteristics

• Imminence: Long-term. The incident poses a long-term strategic concern regarding the normalization of synthetic media use by government institutions.
• Autonomy: Human-controlled. AI tools were directly controlled and prompted by human operators to execute specific image modifications.
• Novelty: Evolved capability. Demonstrates an evolved capability where consumer-grade generative AI is utilized by state actors for domestic political messaging.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or kinetic assets were targeted or affected by this incident.
• Information security: Minor. Official government dissemination of manipulated media on social media represents a minor information integrity issue but lacks foreign adversary involvement.
• Sovereignty: Minor. The doctored photograph could hinder the Justice Department's prosecution of the protester, representing a minor impact on judicial government operations.
• Economic security: Negligible. No strategic economic assets, financial systems, or technological competitive advantages were compromised.
• Societal stability: Minor. The manipulation involved darkening a Black protester's skin and could prejudice her right to a fair trial, representing a minor threat to civil rights.