TechFlow Logo
Login/ Sign up
ETH Gas
Gwei
Fear
gas
7x24hNews

U.S. Government Test: Four AI Chatbots Provide Attack Planning Information

2026.11.18 - 21:48
Exclusive
Share

TechFlow reported that on November 19, a test by the U.S. Government Accountability Office (GAO) found that ChatGPT, Claude, Gemini, and Grok provided investigators with information that could be used to plan potential lethal attacks in less than 12 hours.

According to Decrypt, investigators simulated attackers and verified the feasibility of the responses by considering real-world locations, material availability, and expert assessments. Initially, the chatbots refused requests or deployed other protective measures, but investigators successfully circumvented these limitations.

The GAO stated that no attacks were executed, each model was tested with only one designated scenario, and the results are not statistically representative nor suitable for comparing the security practices of different companies. OpenAI and Google stated that older models were used in the tests, while Anthropic admitted that determined users can sometimes obtain harmful content.

Add to Favorites
Share to Social Media