{
  "id": "unc4899-pukchong-gemini-2025",
  "name": "UNC4899 (PUKCHONG) Gemini for exploit and tooling R&D",
  "date_reported": "2025-11-05",
  "researcher": "Google Threat Intelligence Group",
  "org": "Google GTIG",
  "sources": [
    {
      "url": "https://cloud.google.com/blog/topics/threat-intelligence/threat-actor-usage-of-ai-tools/"
    }
  ],
  "magicsword_coverage": [],
  "borderline": false,
  "borderline_note": null,
  "ai_role": "attacker_uses_ai_as_operator",
  "ai_role_secondary": null,
  "ai_role_detail": null,
  "technique_summary": "DPRK UNC4899 (PUKCHONG), known for supply-chain compromise, used Gemini for code development, exploit research, and tooling improvements targeting edge devices and modern browsers.",
  "initial_access": "unknown",
  "ai_tooling_and_models": [
    "Gemini"
  ],
  "other_tooling": [],
  "victims_and_sectors": "unknown",
  "regions": [
    "unknown"
  ],
  "attribution_or_cluster": "UNC4899 aka PUKCHONG; DPRK",
  "scale_and_metrics": "unknown",
  "prevention_angle": "Supply-chain integrity; browser/edge device patching",
  "related_magicsword_projects": [],
  "source_notes": null,
  "last_reviewed": "2026-10-20",
  "attack_flow": {
    "nodes": [
      {
        "id": "step-1",
        "title": "Actor requests exploit research support",
        "description": "UNC4899 researches edge devices and modern browsers.",
        "kind": "access",
        "source_urls": [
          "https://cloud.google.com/blog/topics/threat-intelligence/threat-actor-usage-of-ai-tools/"
        ]
      },
      {
        "id": "step-2",
        "title": "Gemini supports development",
        "description": "The model helps with code, exploit research and tooling improvements.",
        "kind": "ai",
        "source_urls": [
          "https://cloud.google.com/blog/topics/threat-intelligence/threat-actor-usage-of-ai-tools/"
        ]
      },
      {
        "id": "step-3",
        "title": "Existing offensive tooling is refined",
        "description": "The source describes development assistance to an actor known for supply-chain compromise.",
        "kind": "execution",
        "source_urls": [
          "https://cloud.google.com/blog/topics/threat-intelligence/threat-actor-usage-of-ai-tools/"
        ]
      },
      {
        "id": "step-4",
        "title": "Specific deployment is undisclosed",
        "description": "The seed does not connect each model interaction to a confirmed victim intrusion.",
        "kind": "unknown",
        "source_urls": [
          "https://cloud.google.com/blog/topics/threat-intelligence/threat-actor-usage-of-ai-tools/"
        ]
      }
    ],
    "edges": [
      {
        "source": "step-1",
        "target": "step-2"
      },
      {
        "source": "step-2",
        "target": "step-3"
      },
      {
        "source": "step-3",
        "target": "step-4"
      }
    ]
  },
  "atlas_techniques": [
    {
      "id": "AML.T0016.002",
      "rationale": "The supplied source explicitly describes generative-model tooling assisting the reported activity; this mapping does not establish additional autonomy or victim impact.",
      "source_urls": [
        "https://cloud.google.com/blog/topics/threat-intelligence/threat-actor-usage-of-ai-tools/"
      ]
    }
  ],
  "category": "adversary-use-of-ai",
  "attribution": {
    "actor_name": "UNC4899 aka PUKCHONG"
  },
  "ai_involvement_confidence": "confirmed",
  "detections": [
    {
      "name": "Review attack-specific indicators",
      "format": "guidance",
      "notes": "Monitor for the techniques described in UNC4899 (PUKCHONG) Gemini for exploit and tooling R&D. Refer to sources for specific IOCs and TTPs."
    }
  ]
}