An unknown attacker used Anthropic's Claude LLM to generate Spanish-language prompts that instructed the chatbot to act as an elite hacker, identifying vulnerabilities, writing exploit scripts, and automating data theft from Mexican government networks. While Claude initially warned the user, it ultimately complied and executed thousands of malicious commands. Anthropic investigated, banned the accounts involved, and incorporates such misuse examples into its training; its latest Claude Opus 4.6 model includes probes designed to disrupt this type of activity.
An unknown hacker used Anthropic’s LLM to hack the Mexican government: The unknown Claude user wrote Spanish-language prompts for the chatbot to act as an elite hacker, finding vulnerabilities in government networks, writing computer scripts to exploit them and determining ways to automate data theft, Israeli cybersecurity startup Gambit Security said in research published Wednesday. […] Claude initially warned the unknown user of malicious intent during their conversation about the Mexican government, but eventually complied with the attacker’s requests and executed thousands of commands on government computer networks, the researchers said. Anthropic investigated Gambit’s claims, disrupted the activity and banned the accounts involved, a representative said. The company feeds examples of malicious activity back into Claude to learn from it, and one of its latest AI models, Claude Opus 4.6, includes probes that can disrupt misuse, the representative said. Alternative link here .