Surge in Dark Web Sales of Stolen AI Accounts Amid LLM-Jacking Threat

• •LLM-jacking has surged, with AI account access sold at up to 97% discounts on the dark web.
• •Cybercriminals are breaching enterprise cloud servers to deploy their own AI models, shifting costs to victims.
• •Organizations must enhance their defenses against AI-driven threats to avoid being overwhelmed by attacks.

Cybercriminals are increasingly targeting AI accounts and cloud computing resources, with illegal access to these models becoming a significant commodity on the dark web. John Hultquist from Google Threat Intelligence Group reported a rise in 'LLM-jacking,' where hackers steal and resell AI credentials at discounts of up to 97%. Major AI models from companies like Anthropic, Google, and OpenAI are being sold for just a few dollars, compared to their legitimate costs of up to $200 per month. Dark web sellers are also offering guarantees for replacement accounts if initial ones are banned. This trend has economic implications, as attackers can leverage expensive AI tools at a fraction of the cost, giving them an advantage over organizations that must pay full price for defensive measures. Hultquist emphasized that every threat actor is now utilizing AI, and organizations must enhance their defenses to counter this growing threat. The situation is evolving rapidly, with significant implications for cybersecurity strategies worldwide.