THREATNOIR
Subscribe
Back to Feed
AI SecurityApr 10, 2026

Chat, I've changed my mind. We have some problems in the AI department. It turns out someone com...

Mexican government extensively compromised using Claude and ChatGPT.

Read at source
Vendor Watch

Run Anthropic?

Get an email when a reviewed story names Anthropic, usually within the hour.

Free. Your list stays private and never appears in a subject line. One click stops it. How Vendor Watch works

Summary

A report details how the Mexican government was compromised at scale using nothing but Claude and ChatGPT. The breach appears to involve unauthorized access to sensitive government systems and data, leveraging publicly available AI tools. The full technical details are provided in an accompanying paper.

Entities

Claude (product)ChatGPT (product)Anthropic (vendor)OpenAI (vendor)