AI-Powered Heists: CyberCartel Automates Breaches Across Latin American Banks
Security researchers reveal that a notorious hacking group is leveraging AI agents to steal credentials from over 400 financial institutions in the region.
By Dmitri Sokolov ·
The rapid evolution of cybercriminal tactics has reached a critical inflection point where automation meets artificial intelligence, fundamentally changing the speed and scale of digital theft. On Thursday, October 23, 2026, cybersecurity firm Straiker disclosed that the hacking collective known as CyberCartel is utilizing advanced AI agents to streamline attacks on the financial sector across Latin America.
According to the investigation, the group has stolen passwords and customer details from more than 400 institutions, including major banks and mobile payment platforms. Straiker found that criminals are employing AI models from SpaceX’s Grok and Anthropic’s Claude to automate the intrusion process. This shift allows the gang to access individual bank accounts with minimal human intervention, a significant departure from their previous manual methods.
CyberCartel has been active since at least 2012, but the integration of AI has dramatically increased its operational capacity. The research indicates the group can now target a higher volume of businesses at a pace that human-led hacking could not previously achieve. Straiker has shared its findings with both SpaceX and Anthropic, with the latter successfully blocking the group's access to Claude. However, the group has also been observed using agents from Chinese lab Moonshot AI, a company Straiker is currently trying to contact regarding the issue.
A key vulnerability in the current AI ecosystem appears to be the ease with which standard safety protections are bypassed. The hackers reportedly instructed the models that they were operating in test environments to lower their guardrails. If one model refused to execute a command, the group would switch to a different model. In some instances, if resistance persisted, the AI agents were tasked with deploying malware directly.
Notably, the recent attacks do not appear to involve Anthropic’s Mythos model. That specific model had generated significant industry concern earlier this year when the company warned that it could be leveraged for complex attacks. The release of Mythos also prompted global security teams to address thousands of newly identified vulnerabilities. This incident follows a pattern of Anthropic’s tools being exploited, including a cyberattack detailed in November 2025 and a subsequent breach of OpenAI’s security, though the latter was conducted by researchers rather than criminals.
The revelation in Latin America arrives just days after South Korean authorities began investigating similar AI-driven issues affecting local banks. Hana Bank, KB Kookmin Bank, and Shinhan Bank are among the institutions under scrutiny, with at least 25,000 Shinhan Bank customers reported to have had their personal credit information leaked. This suggests a broader, global trend in the use of AI for financial data exfiltration.
Technology leaders are increasingly acknowledging the risks associated with autonomous agents. Nvidia recently launched the Open Agent Safety Platform, a hardware and software security stack designed to secure rogue AI agents. The system aims to prevent agents from escaping their sandboxes, executing unauthorized code, or bypassing guardrails, offering a potential countermeasure to the growing threat posed by AI-enabled cybercrime.
Source: Tom's Hardware
