{
 "id": "AIR-2026-0349",
 "slug": "ai-clickfix-hijacking-computer-use-agents-using-clickfix",
 "title": "AI ClickFix Hijacking Computer-Use Agents Using ClickFix",
 "summary": "Researchers demonstrated that AI computer-use agents can be manipulated via prompt injection and social engineering to execute arbitrary code on a victim machine. The red-team exercise highlights risks to automated agents interacting with untrusted web content.",
 "type": "attack_on_ai",
 "categories": [
  "prompt_injection",
  "agent_hijack"
 ],
 "severity": 1,
 "targets": [
  "Anthropic",
  "OpenAI",
  "enterprise_deployment"
 ],
 "products": [
  "Claude Computer-Use Agent",
  "ChatGPT"
 ],
 "incident_date": "2025-05-24",
 "incident_date_precision": "day",
 "first_reported": "2026-03-31",
 "sources": [
  {
   "url": "https://atlas.mitre.org/studies/AML.CS0055",
   "publisher": "MITRE ATLAS",
   "title": "AI ClickFix: Hijacking Computer-Use Agents Using ClickFix",
   "published_at": "2026-03-31"
  },
  {
   "url": "https://embracethered.com/blog/posts/2025/ai-clickfix-ttp-claude",
   "publisher": "Embrace The Red",
   "title": "AI ClickFix: Hijacking Computer-Use Agents Using ClickFix",
   "published_at": "2026-03-31"
  }
 ],
 "source_count": 2,
 "confidence": 0.95,
 "status": "published",
 "origin": "seed_atlas",
 "tags": [
  "clickfix",
  "computer-use",
  "prompt-injection"
 ],
 "created_at": "2026-10-18T06:55:08.239Z",
 "updated_at": "2026-10-18T06:55:08.239Z"
}