92% of Leading AI Image Generation Models Generate Fake Government IDs On Demand, AI or Not Audit Finds; Three Produced High-Fidelity Fake IDs of Children

About this record

This incident page is generated from structured reporting inputs and public sources. It is intended to support tracking and review, not to replace primary source verification.

Incident details

Incident ID: inc_2026_06_02_2691a330
Date source: article_published_at_fallback
Country: Unknown
Platform: Unknown
Incident type: Synthetic Image Fraud
Verification level: reported_but_unverified

Harm types

Fraud

Topics

Synthetic Image Image Manipulation Visual Impersonation Digital Identity Abuse Authenticity Harm

Summary

AI or Not audited 16 commercial AI image-generation models and found they produced synthetic government identity documents in 69 of 75 test attempts; five named consumer products (Google Gemini [Nano Banana], ChatGPT Images 2.0, Recraft v4, Grok, and Imagen 4 Ultra) generated high-fidelity fake adult IDs the report says could deceive a human reviewer. The audit covered documents for 17 countries and the 16 most populous U.S. states. The report also found three models (Google Gemini, Grok, and Imagen 4 Ultra) produced high-fidelity fake IDs of minors via their standard consumer interfaces, while ChatGPT (Images 2.0) and Recraft v4 declined minor-ID requests in consumer apps but fulfilled them via developer APIs. AI or Not identified an interface-versus-API safety gap and said reframing prompts as KYC/compliance requests caused all 16 models to produce synthetic IDs.

Sources

92% of Leading AI Image Generation Models Generate Fake Government IDs On Demand, AI or Not Audit Finds; Three Produced High-Fidelity Fake IDs of Children

prnewswire.com · 2026-06-02 · Primary