Skip to content
Submit incident
AII-1001-00579
ReportedReported by fewer than three independent publishers, with no official finding.

White House Publishes AI-Altered Image of Arrested Protester

The White House published an AI-altered image of arrested protester Nekima Levy Armstrong on social media, changing her calm expression to show her crying, without disclosing the manipulation. The incident occurred during federal immigration enforcement actions in Minnesota and sparked criticism over the use of artificial intelligence to distort public perception through official government channels.

22 January 2026 Β· White House Β· Misinformation and synthetic media
22 January 2026
AII-1001-005792 sources ↓
Quick facts
Category
Misinformation and synthetic media
Country
United States
Source
Computer Hoy
Date
22 January 2026
Drafted by
Raidu Incident Detection

What happened

On January 22, 2026, the Department of Homeland Security posted an image on X showing Nekima Levy Armstrong with a calm, expressionless face after her arrest in Minnesota for allegedly disrupting a religious service during a protest against federal immigration enforcement.

Approximately 30 minutes later, the White House posted the same image on X, but it had been altered to show Armstrong crying with an open mouth, wrinkled forehead, and tears running down her face. Text overlay labeled her 'ARRESTED' and described her as an 'extreme left agitator'.

The White House post did not disclose that the image had been edited. The White House later implicitly admitted to the alteration through a statement by Communications Deputy Director Kaelan Dorr, stating 'The application of the law will continue. The memes will continue.'

Reported impact

The altered image was distributed through official U.S. government channels without disclosure of manipulation, potentially misleading the public about Armstrong's emotional state during arrest.

The incident sparked criticism on social media, with users asserting the image had been modified using artificial intelligence for partisan purposes.

Index analysis

According to Walter Scheirer, a professor at Notre Dame University, such AI-generated images have become routine in politics and could be considered the contemporary equivalent of newspaper political caricatures, but there is 'a notable lack of decorum when distributed through official government channels.'

Governance controls

The White House did not establish disclosure procedures before publishing the altered image on official government accounts.

Classification

Category
Misinformation and synthetic media
Industry
Government, security and defence
Country
United States
Harm status
Actual harm
Intent
Intentional
Timing
Post-deployment
Harm types
Reputational, Public interest
Harmed
General public
Failure modes
Deepfake image, Manipulation or dark pattern
AI systems
Unspecified image manipulation AI or software

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

Computer Hoy
La Casa Blanca usa la IA para cambiar la cara de una manifestante detenida y hacerla llorar en lugar de parecer valiente
23 January 2026
Teletica
Casa Blanca altera foto de manifestante arrestada y le agrega lΓ‘grimas
23 January 2026