AI-Powered Fraud

{
  "id": "fraud",
  "numericId": null,
  "path": "/knowledge-base/risks/fraud/",
  "filePath": "knowledge-base/risks/fraud.mdx",
  "title": "AI-Powered Fraud",
  "quality": 69,
  "readerImportance": 58.3,
  "researchImportance": 69,
  "tacticalValue": 72,
  "contentFormat": "article",
  "tractability": null,
  "neglectedness": null,
  "uncertainty": null,
  "causalLevel": "outcome",
  "lastUpdated": "2026-03-13",
  "dateCreated": "2026-02-15",
  "llmSummary": "Comprehensive reference on AI-enabled fraud covering technical pipelines, case studies, and countermeasures, anchored by FBI IC3 2024 data (\\$16.6B total reported losses, +33% YoY); critically notes that AI-specific share of losses is not disaggregated in official statistics and removes unverified projections. Detection technology shows severe generalization gaps (~50% AUC drop on real-world deepfakes vs. benchmarks), with human detection barely above chance (~55-60%).",
  "description": "AI enables automated fraud at scale through voice cloning, personalized phishing, and deepfake video calls. FBI-reported losses reached \\$16.6B in 2024, a 33% increase from 2023.",
  "ratings": {
    "focus": 8.5,
    "novelty": 4.5,
    "rigor": 8,
    "completeness": 8.5,
    "concreteness": 8,
    "actionability": 6.5,
    "objectivity": 8.5
  },
  "category": "risks",
  "subcategory": "misuse",
  "clusters": [
    "cyber",
    "ai-safety"
  ],
  "metrics": {
    "wordCount": 4516,
    "tableCount": 6,
    "diagramCount": 0,
    "internalLinks": 10,
    "externalLinks": 23,
    "footnoteCount": 0,
    "bulletRatio": 0.24,
    "sectionCount": 30,
    "hasOverview": true,
    "structuralScore": 14
  },
  "suggestedQuality": 93,
  "updateFrequency": 21,
  "evergreen": true,
  "wordCount": 4516,
  "unconvertedLinks": [
    {
      "text": "\"Human performance in detecting deepfakes: A systematic review and meta-analysis of 56 papers\"",
      "url": "https://www.sciencedirect.com/science/article/pii/S2451958824001714",
      "resourceId": "5c1ad27ec9acc6f4",
      "resourceTitle": "Human performance in detecting deepfakes: A systematic review and meta-analysis"
    },
    {
      "text": "\"Deepfake-Eval-2024: A Multi-Modal In-the-Wild Benchmark of Deepfakes Circulated in 2024\"",
      "url": "https://arxiv.org/html/2503.02857v2",
      "resourceId": "f39c2cc4c0f303cc",
      "resourceTitle": "Deepfake-Eval-2024 benchmark"
    },
    {
      "text": "EU AI Act Article 50",
      "url": "https://artificialintelligenceact.eu/article/50/",
      "resourceId": "44e36a446a9f4de6",
      "resourceTitle": "EU AI Act Article 50"
    }
  ],
  "unconvertedLinkCount": 3,
  "convertedLinkCount": 0,
  "backlinkCount": 6,
  "hallucinationRisk": {
    "level": "medium",
    "score": 40,
    "factors": [
      "no-citations",
      "high-rigor"
    ]
  },
  "entityType": "risk",
  "redundancy": {
    "maxSimilarity": 19,
    "similarPages": [
      {
        "id": "fraud-sophistication-curve",
        "title": "Fraud Sophistication Curve Model",
        "path": "/knowledge-base/models/fraud-sophistication-curve/",
        "similarity": 19
      },
      {
        "id": "epistemic-security",
        "title": "AI-Era Epistemic Security",
        "path": "/knowledge-base/responses/epistemic-security/",
        "similarity": 18
      },
      {
        "id": "disinformation",
        "title": "Disinformation",
        "path": "/knowledge-base/risks/disinformation/",
        "similarity": 18
      },
      {
        "id": "agentic-ai",
        "title": "Agentic AI",
        "path": "/knowledge-base/capabilities/agentic-ai/",
        "similarity": 17
      },
      {
        "id": "authoritarian-tools-diffusion",
        "title": "Authoritarian Tools Diffusion Model",
        "path": "/knowledge-base/models/authoritarian-tools-diffusion/",
        "similarity": 17
      }
    ]
  },
  "coverage": {
    "passing": 6,
    "total": 13,
    "targets": {
      "tables": 18,
      "diagrams": 2,
      "internalLinks": 36,
      "externalLinks": 23,
      "footnotes": 14,
      "references": 14
    },
    "actuals": {
      "tables": 6,
      "diagrams": 0,
      "internalLinks": 10,
      "externalLinks": 23,
      "footnotes": 0,
      "references": 23,
      "quotesWithQuotes": 0,
      "quotesTotal": 0,
      "accuracyChecked": 0,
      "accuracyTotal": 0
    },
    "items": {
      "llmSummary": "green",
      "schedule": "green",
      "entity": "green",
      "editHistory": "red",
      "overview": "green",
      "tables": "amber",
      "diagrams": "red",
      "internalLinks": "amber",
      "externalLinks": "green",
      "footnotes": "red",
      "references": "green",
      "quotes": "red",
      "accuracy": "red"
    },
    "ratingsString": "N:4.5 R:8 A:6.5 C:8.5"
  },
  "readerRank": 242,
  "researchRank": 159,
  "recommendedScore": 189.01
}

{
  "lesswrong": "https://www.lesswrong.com/tag/ai-misuse"
}