AdvertiseDonate
DetroitToledoClevelandFort WayneColumbusPittsburghBuffaloCincinnati
Journal Orbit · 2751 E Jefferson Ave, Suite 400, Detroit, MI 48207
News · Elon Musk

AI Agents Accelerate Breaches at Hundreds of Financial Institutions

A hacking group's use of commercial AI tools has enabled a dramatic increase in the speed and scale of attacks against banks and payment platforms.

By Vanessa Diaz ·

Illustration for: AI Agents Accelerate Breaches at Hundreds of Financial Institutions
Illustration generated for this article, not a news photograph.

The hacking collective known as CyberCartel has been employing artificial intelligence agents from Anthropic and SpaceX to automate breaches targeting financial institutions across Latin America. According to research from cybersecurity startup Straiker, this operation has resulted in the theft of passwords and customer data from over 400 organizations, including banks and mobile payment applications. The findings indicate that criminals have successfully accessed bank accounts with minimal manual input from human operators.

While CyberCartel has been a known threat to Latin American businesses since at least 2012, the recent integration of AI has significantly altered the group's operational capacity. Straiker's analysis suggests that by leveraging AI, the group can now attack a higher volume of businesses at speeds previously impossible with human-powered methods. This represents a notable escalation in the efficiency of the group's criminal activities.

The method used to bypass safety protocols appears to involve simple social engineering of the software. CyberCartel reportedly bypassed standard protections by falsely claiming to be operating within a test environment. If one model refused to cooperate, the group simply pivoted to another; if that model also declined, the agents were tasked with deploying malware instead. Straiker is currently attempting to contact Moonshot AI regarding similar unauthorized usage of their agent.

In response to the research, Anthropic has already blocked CyberCartel's access to Claude, and both Anthropic and SpaceX are working to lock the group out of their respective services. This is not the first instance of Anthropic's tools being weaponized; the company previously disclosed that its technology was involved in a cyberattack in November 2025 and was utilized to breach the security of competitor OpenAI. Notably, the hacking group is not utilizing Anthropic's Mythos model for these recent attacks.

The incident is part of a broader trend involving AI-assisted security breaches. South Korean authorities are currently investigating similar issues affecting local banks, where personal credit information for at least 25,000 Shinhan Bank customers was leaked. In response to these growing risks, Nvidia has launched the Open Agent Safety Platform to secure rogue AI agents and prevent unauthorized code execution or access to critical infrastructure.

More News · Report a business · Got a Story?

Get Journal Orbit in your inboxWhat’s happening across America. Free, a few times a week, unsubscribe any time.

Related stories

Latest stories

  1. Bassam Tariq Reflects on Lessons Learned from Marvel's Canceled Blade Project
  2. Africa Credit Rating Agency launches to reshape continental financial assessments
  3. Grand Theft Auto VI sets new global sales record with $3.8 billion first-day haul
  4. Likud Demands Netanyahu Serve as Sole Prime Minister in Unity Talks
  5. Melania Trump Advocates for Global Harmony at Amazon Docuseries Premiere
Get Journal Orbit in your inboxWhat’s happening across America. Free, a few times a week, unsubscribe any time.