AUTO-UPDATED

Hackers hijack AI accounts and servers to fuel a booming cyber crime economy

Cybercriminals are increasingly targeting corporate AI accounts through "LLMjacking," using stolen API keys and session tokens to exploit expensive compute resources for illicit, large-scale operations.

Key Points

  • Hackers use infostealer malware to siphon credentials, allowing them to resell access to major AI platforms like OpenAI, Anthropic, and Google at significant discounts.
  • CrowdStrike reported an 89% increase in AI-related adversary activity between July 2025 and June 2026.
  • A September 2026 analysis by Okta identified hundreds of unexpired AI service tokens within a single 7 GB dump of stolen logs.
  • Autonomous AI-agent campaigns compromised infrastructure across 395 organizations in 48 countries during September 2026.
  • Attackers often evade detection by throttling usage to mimic legitimate traffic patterns, making unauthorized activity difficult for traditional security systems to identify.

Why it Matters

This trend forces companies to face massive, unauthorized compute bills while risking the exposure of sensitive data processed through hijacked AI models. As these attacks become more sophisticated, organizations must adopt more rigorous credential management and monitoring to protect their infrastructure from evolving cyber threats.
Crypto Briefing Published by Diego Almada Lopez
Read original