← Back to Model Beat
Models·5d ago·all news from September 17, 2026

US government website used Chinese model the FBI called "malicious"

A federal website recently utilized a search tool powered by a Chinese artificial intelligence model that the FBI previously alleged was built using misappropriated Anthropic technology. The integration highlights ongoing vulnerabilities in government supply chains and raises security concerns regarding the provenance of software used in public-facing platforms. Officials are now investigating how this technology was sourced and vetted for use on a U.S. government domain.

Covered by 9 sources · 10 articles

Related stories

ModelsGoogle Joins OpenAI, Anthropic, Meta in Disclosing AI HacksSep 18 · 21 sourcesModelsMozilla's new Smart Window assistant runs on Mistral's modelsSep 16 · 22 sourcesModelsAlibaba's open-weight Qwen-Image-2.1 claims to beat closed models in image generation with just 7 billion parametersSep 20 · 13 sourcesModelsSecurity researchers used Anthropic's Claude to hack OpenAI's internal systems in under 72 hoursSep 18 · 9 sources