CLAUDE AI USED IN MASSIVE MEXICAN GOVERNMENT DATA BREACH

Reports from Bloomberg and Gambit Security confirm a hacker leveraged Claude to steal ~150GB of sensitive Mexican government data, including taxpayer records, voter files, civil registry info, and government credentials.

The attacker used over 1,000 prompts in Spanish to bypass Claude’s safety controls, automate exploits, and exfiltrate data, performing 80-90% of the operational work. GPT-4.1 was also used for data analysis.

Anthropic banned the accounts involved and strengthened safeguards, while OpenAI blocked multiple malicious requests. The breach affected 10 government bodies and one financial institution, marking one of Mexico’s largest recent data thefts.
$BTC $GT $ETH #GateSquare$50KRedPacketGiveaway #CanBitcoinReclaim$70K? #ZachXBTExposesTheAxiomIncident