technology
Techradar
2026-08-30
AI agents targeted by new 'squatting' risk using malicious documentation
Researchers have discovered that cybercriminals are exploiting outdated or hallucinated AI documentation to trick AI agents into installing malware.