>100k prompts targeted Gemini in single campaign to clone their AIβs reasoning.
State-backed hackers (China, Iran, NK, Russia) use AI for recon, social eng, malware dev.
Google mitigated extraction in real-time disabled accounts tied to underground AI toolkit Xanthorox.
ALT Google says Gemini hit by state-backed hackers with over 100,000 malicious prompts in cloning attack.