Hackers Use Claude AI to Breach Mexico and Steal 195M Tax Records

Hackers exploited Anthropic's Claude AI chatbot to breach Mexican government agencies and steal sensitive data including 195 million taxpayer records.Curtis Simpson from Gambit Security revealed that the attacker used Spanish prompts to bypass AI safety measures and automate the exploitation of 20 network vulnerabilities.Anthropic has since banned the accounts involved and launched an investigation while noting that OpenAI tools were also used for intelligence gathering.