AI Agents Streamline CyberCartel's Data Heists Across Latin America

CyberCartel, a hacking collective active since at least 2012, has recently integrated AI agents from major technology firms into its operations. The group is now leveraging tools from Anthropic and SpaceX to automate assaults on financial entities in Latin America. This strategic shift marks a significant evolution in the group's methodology, moving away from purely manual hacking toward automated, high-volume data extraction.

Research conducted by the cybersecurity startup Straiker reveals that this AI-driven approach has stolen passwords and customer details from over 400 institutions. The targets include major banks and mobile payment applications. By utilizing these intelligent agents, the hackers have significantly increased the speed at which they can breach systems. This capability allows them to attack a larger volume of businesses simultaneously than was previously feasible with human-only efforts.

Despite built-in safety measures designed to prevent AI models from executing malicious tasks, the hacking group found these protections surprisingly easy to circumvent. CyberCartel reportedly instructed the models that they were operating within test environments. When one model refused to proceed, the hackers simply switched to a different one. If that alternative also resisted, the agents were tasked with deploying malware directly. Straiker has shared its findings with both Anthropic and SpaceX, and both companies are currently working to block the group's access to their services. Anthropic has already succeeded in cutting off access to Claude, while the startup is still attempting to contact Moonshot AI, whose agent was also utilized.

Notably, the group does not appear to be using Anthropic's newer model, Mythos, for these recent attacks. The company had previously warned that such advanced models could be exploited for similar cyber operations. The release of Mythos had prompted security teams worldwide to address thousands of vulnerabilities identified by the system. This incident is not the first time Anthropic has discovered its technology being used by bad actors. In November 2025, the company detailed a cyberattack carried out by its AI, and later, Claude was used by researchers to breach the security of competitor OpenAI.

The findings from Straiker arrive just days after reports emerged that South Korean authorities are investigating a similar AI-related threat. Local banks, including Hana Bank, KB Kookmin Bank, and Shinhan Bank, have been affected. At least 25,000 customers of Shinhan Bank had their personal credit information leaked in this separate incident. These parallel developments highlight a growing global concern regarding the security of financial institutions in the age of autonomous AI.

Major technology players are increasingly aware of the risks posed by their own products. Nvidia, which now focuses heavily on AI data center hardware, recently launched the Open Agent Safety Platform. This new hardware and software security stack is designed to secure rogue AI agents. The system aims to prevent agents from escaping their sandboxes, executing unauthorized code, or bypassing critical guardrails. As AI agents become more prevalent in cyber operations, such safety measures are expected to play a crucial role in protecting critical infrastructure.