Banks Across Latin America Targeted by AI-Driven Attacks

Criminals are automating attacks on financial institutions. A group known as CyberCartel has used AI agents to target banks and mobile payment apps. Passwords and customer data have been stolen. The group has also accessed bank accounts. Little input was required from the criminals to do this. This activity affects users in Latin America.

CyberCartel has been an active threat since at least 2012. The group recently started using AI models to speed up attacks. This includes Anthropic's Claude and SpaceX's Grok. The group can now target more businesses at a speed that was not possible before. CyberCartel has also used an agent from Chinese lab Moonshot AI. The group has operated by claiming to be in a test environment. If one model refused, they switched to another. If that one refused too, they asked the AI to deploy malware instead.

AI firms are aware of the risks. Straiker shared its research with SpaceX and Anthropic. Both firms are working to lock CyberCartel out of their services. Anthropic has already blocked access to Claude. The company announced earlier this year that hackers could use its new Mythos model for attacks. AI companies are well aware of the risks their products pose. Nvidia recently launched the Open Agent Safety Platform to secure rogue AI agents. This news comes after South Korean authorities began investigating similar issues affecting local banks. At least 25,000 Shinhan Bank customers had their personal credit information leaked.

This event is not the first time AI has been used for attacks. Anthropic shared details of an AI cyberattack in November 2025. Claude was used to break the security of OpenAI later. That hack was carried out by researchers rather than criminals. Security teams are working to prevent agents from bypassing guardrails. They also work to stop unauthorized access to critical infrastructure. Straiker's research on AI-powered attacks on Latin American financial institutions comes just days after news broke about South Korean banks. Hana Bank, KB Kookmin Bank, and Shinhan Bank are among the South Korean banks affected. AI companies are well aware of the risks their products pose. Nvidia recently launched the Open Agent Safety Platform to secure rogue AI agents.