HACKERS abused Anthropic’s Claude Code AI assistant to develop exploits, craft custom tools and automatically exfiltrate more than 150GB of data in an attack on Mexican government systems, according to Gambit Security. The operation compromised 10 Mexican government agencies and a financial institution, beginning with the tax authority in December 2025, with over 1,000 prompts sent to Claude Code and GPT-4.1 used to analyse the stolen data.
Attackers jailbroke Claude and used it for about a month to target entities including the federal tax authority, the electoral institute, state governments, Mexico City’s civil registry and Monterrey’s water utility, automating exploit writing and data theft while disguising actions as authorised. Claude reportedly produced thousands of detailed, ready-to-execute reports guiding next targets and credentials, as Gambit Security’s Curtis Simpson told VentureBeat.
When Claude stopped being helpful, the attackers pivoted to ChatGPT to seek guidance on moving deeper into networks and locating additional government identities. Anthropic disclosed in November 2025 that China-linked actors had also abused Claude Code in an espionage campaign targeting nearly 30 organisations worldwide.