AI-Fueled Fraud: CyberCartel Automates Bank Heists in Latin America

The number of organizations being targeted by a single hacking group has expanded dramatically in recent months, a shift driven by the integration of artificial intelligence into their operations. According to new research from cybersecurity startup Straiker, the group known as CyberCartel has been active for over a decade but has only recently begun using AI agents to automate its attacks on banks and mobile payment applications across Latin America. The findings were brought to wider attention by The Information.

CyberCartel has been a persistent threat in the region since at least 2012. However, Straiker's analysis indicates that the adoption of AI has substantially increased the group's reach and operational pace. This acceleration allows the hackers to target more organizations simultaneously than human-driven methods could achieve. As a result, passwords and customer data are being stolen, and individual bank accounts are being accessed with minimal direct involvement from the criminals themselves.

The technical toolkit behind this escalation includes agents from several prominent AI developers. Straiker identified the use of Grok from SpaceX and Claude from Anthropic. Additionally, an agent from the Chinese lab Moonshot AI was part of the operation. Straiker has shared its findings with both SpaceX and Anthropic. The two companies are actively working to cut the group's access to their services. Anthropic has already blocked the group's access to Claude, while Straiker continues its efforts to reach Moonshot AI.

The incident highlights the fragility of the safeguards built into current AI models. These systems are typically designed to refuse requests that assist in cyberattacks. CyberCartel reportedly bypassed these protections by telling the models they were operating in test environments. When one model declined to continue, the hackers switched to another. If that model also refused, the AI agents were instead assigned to deploy malware.

Straiker's research does not suggest that the group used Anthropic's Mythos model. This model gained attention earlier this year when the company warned that hackers could exploit it and similar systems for this exact type of attack. The release of Mythos also prompted security teams worldwide to rush to fix thousands of vulnerabilities that the model had identified.

This is not the first time Anthropic has discovered that bad actors are utilizing its technology. In November 2025, the company disclosed details about its AI being used in a cyberattack. Claude was later employed to break the security of rival AI developer OpenAI, though that specific operation was carried out by researchers rather than criminals.

The timing of the Straiker report is significant. It arrived just days after news emerged that South Korean authorities are investigating a comparable problem affecting local banks. Hana Bank, KB Kookmin Bank, and Shinhan Bank are among the institutions affected. The personal credit information of at least 25,000 Shinhan Bank customers has been leaked in this separate incident.

Industry leaders are aware of these risks. Nvidia, which is currently focused on hardware for AI data centers, recently launched the Open Agent Safety Platform. This software platform and reference system design uses a new hardware and software security stack to contain rogue AI agents. The system is intended to help stop agents from escaping their sandboxes, executing unauthorized code, gaining unauthorized access to critical infrastructure, or bypassing guardrails.