{
  "id": "sentinellabs-uncovers-malterminal-gpt-4-powered-2025-09",
  "name": "SentinelLABS uncovers MalTerminal, GPT-4-powered malware, and hunts LLM malware by embedded keys and prompts",
  "date_reported": "2025-09-19",
  "researcher": "unknown",
  "org": "SentinelLABS",
  "sources": [
    {
      "url": "https://www.sentinelone.com/labs/prompts-as-code-embedded-keys-the-hunt-for-llm-enabled-malware/",
      "publisher": "SentinelLABS"
    },
    {
      "url": "https://thehackernews.com/2025/09/researchers-uncover-gpt-4-powered.html"
    },
    {
      "url": "https://securityaffairs.com/182433/malware/researchers-expose-malterminal-an-llm-enabled-malware-pioneer.html"
    }
  ],
  "borderline": false,
  "ai_role": "llm_generated_malware",
  "ai_role_secondary": "ai_credential_theft",
  "technique_summary": "SentinelLABS identified possibly the earliest LLM-enabled malware and showed that embedded API keys and prompts give defenders a reliable hunting handle.",
  "initial_access": "unknown",
  "ai_tooling_and_models": [
    "OpenAI GPT-4 chat completions API (deprecated endpoint)",
    "Hugging Face keys in LameHug"
  ],
  "other_tooling": [],
  "victims_and_sectors": "Multiple sectors",
  "regions": [
    "unknown"
  ],
  "region_notes": "Target geography was not captured by the Notion feed. Actor attribution is not used as a victim location.",
  "attribution_or_cluster": "Unknown author of MalTerminal",
  "scale_and_metrics": "unknown",
  "prevention_angle": "unknown",
  "related_magicsword_projects": [],
  "source_notes": "Imported from the AI Threat Watch Notion feed (an automated daily scan); not yet editorially reviewed. Feed severity: Medium. Feed novelty: New TTP.",
  "tags": [
    "notion-import",
    "malware-tooling-dev",
    "api-key-account-abuse"
  ],
  "category": "adversary-use-of-ai",
  "ai_involvement_confidence": "confirmed"
}