Multi-Agent Safety

{
  "id": "multi-agent",
  "wikiId": "E488",
  "path": "/knowledge-base/responses/multi-agent/",
  "filePath": "knowledge-base/responses/multi-agent.mdx",
  "title": "Multi-Agent Safety",
  "quality": 68,
  "readerImportance": 21,
  "researchImportance": 32.5,
  "tacticalValue": null,
  "contentFormat": "article",
  "causalLevel": null,
  "lastUpdated": "2026-01-29",
  "dateCreated": "2026-02-15",
  "summary": "Multi-agent safety addresses coordination failures, conflict, and collusion risks when AI systems interact. A 2025 report from 50+ researchers identifies seven key risk factors; empirical studies show 35-76% of LLMs exploit coordination incentives, while safe MARL algorithms (MACPO) achieve near-zero constraint violations in benchmarks. Current research investment (\\$5-15M/year) is significantly below single-agent alignment (\\$100M+), despite the AI agents market projected to grow from \\$5.4B (2024) to \\$236B by 2034.",
  "description": "Multi-agent safety research addresses coordination failures, conflict, and collusion risks when multiple AI systems interact.",
  "ratings": {
    "novelty": 6.5,
    "rigor": 7,
    "completeness": 7.5,
    "actionability": 6
  },
  "category": "responses",
  "subcategory": "alignment-deployment",
  "clusters": [
    "ai-safety"
  ],
  "metrics": {
    "wordCount": 3606,
    "tableCount": 17,
    "diagramCount": 1,
    "internalLinks": 19,
    "externalLinks": 31,
    "footnoteCount": 0,
    "bulletRatio": 0.12,
    "sectionCount": 31,
    "hasOverview": true,
    "structuralScore": 15
  },
  "suggestedQuality": 100,
  "updateFrequency": 21,
  "evergreen": true,
  "wordCount": 3606,
  "unconvertedLinks": [
    {
      "text": "2025 survey on safe reinforcement learning",
      "url": "https://arxiv.org/html/2505.17342v1",
      "resourceId": "7ba5b02ca89ba9eb",
      "resourceTitle": "MACPO (Multi-Agent Constrained Policy Optimization)"
    },
    {
      "text": "arXiv:2502.14143",
      "url": "https://arxiv.org/abs/2502.14143",
      "resourceId": "772b3b663b35a67f",
      "resourceTitle": "2025 technical report"
    },
    {
      "text": "Blog post",
      "url": "https://www.cooperativeai.com/post/new-report-multi-agent-risks-from-advanced-ai",
      "resourceId": "05b7759687747dc2",
      "resourceTitle": "Cooperative AI Foundation's taxonomy"
    },
    {
      "text": "arXiv:2505.17342",
      "url": "https://arxiv.org/html/2505.17342v1",
      "resourceId": "7ba5b02ca89ba9eb",
      "resourceTitle": "MACPO (Multi-Agent Constrained Policy Optimization)"
    }
  ],
  "unconvertedLinkCount": 4,
  "convertedLinkCount": 14,
  "backlinkCount": 6,
  "hallucinationRisk": {
    "level": "low",
    "score": 30,
    "factors": [
      "no-citations",
      "high-rigor",
      "conceptual-content"
    ]
  },
  "entityType": "approach",
  "redundancy": {
    "maxSimilarity": 17,
    "similarPages": [
      {
        "id": "reasoning",
        "title": "Reasoning and Planning",
        "path": "/knowledge-base/capabilities/reasoning/",
        "similarity": 17
      },
      {
        "id": "intervention-effectiveness-matrix",
        "title": "Intervention Effectiveness Matrix",
        "path": "/knowledge-base/models/intervention-effectiveness-matrix/",
        "similarity": 17
      },
      {
        "id": "scalable-oversight",
        "title": "Scalable Oversight",
        "path": "/knowledge-base/responses/scalable-oversight/",
        "similarity": 17
      },
      {
        "id": "sharp-left-turn",
        "title": "Sharp Left Turn",
        "path": "/knowledge-base/risks/sharp-left-turn/",
        "similarity": 17
      },
      {
        "id": "treacherous-turn",
        "title": "Treacherous Turn",
        "path": "/knowledge-base/risks/treacherous-turn/",
        "similarity": 17
      }
    ]
  },
  "coverage": {
    "passing": 7,
    "total": 13,
    "targets": {
      "tables": 14,
      "diagrams": 1,
      "internalLinks": 29,
      "externalLinks": 18,
      "footnotes": 11,
      "references": 11
    },
    "actuals": {
      "tables": 17,
      "diagrams": 1,
      "internalLinks": 19,
      "externalLinks": 31,
      "footnotes": 0,
      "references": 7,
      "quotesWithQuotes": 0,
      "quotesTotal": 0,
      "accuracyChecked": 0,
      "accuracyTotal": 0
    },
    "items": {
      "summary": "green",
      "schedule": "green",
      "entity": "green",
      "editHistory": "red",
      "overview": "green",
      "tables": "green",
      "diagrams": "green",
      "internalLinks": "amber",
      "externalLinks": "green",
      "footnotes": "red",
      "references": "amber",
      "quotes": "red",
      "accuracy": "red"
    },
    "ratingsString": "N:6.5 R:7 A:6 C:7.5"
  },
  "readerRank": 514,
  "researchRank": 398,
  "recommendedScore": 160.02
}

{
  "wikipedia": "https://en.wikipedia.org/wiki/Multi-agent_system",
  "wikidata": "https://www.wikidata.org/wiki/Q85786957",
  "grokipedia": "https://grokipedia.com/page/Multi-agent_system"
}