{
  "apiVersion": "v1",
  "methodology": "https://sourcescore.org/methodology/",
  "canonical": "https://sourcescore.org/claims/6fa575eb9df5ac32/",
  "claim": {
    "vertical": "ai-ml",
    "subject": "Anthropic Constitutional AI Harmlessness",
    "predicate": "introduced_in_paper",
    "object": "Bai et al. 2022 — training a helpful and harmless assistant",
    "confidence": 1,
    "sources": [
      {
        "url": "https://arxiv.org/abs/2212.08073",
        "title": "Constitutional AI: Harmlessness from AI Feedback",
        "publisher": "arXiv (Bai, Kadavath, Kundu, Askell, Kernion, Jones, Chen, et al. / Anthropic)",
        "publishedDate": "2022-12-15",
        "accessedDate": "2026-05-16",
        "type": "preprint",
        "excerpt": "As AI systems become more capable, we would like to enlist their help to supervise other AIs. We experiment with methods for training a harmless AI assistant through self-improvement, without any human labels identifying harmful outputs."
      },
      {
        "url": "https://www.anthropic.com/news/claudes-constitution",
        "title": "Claude's Constitution",
        "publisher": "Anthropic",
        "publishedDate": "2023-05-09",
        "accessedDate": "2026-05-16",
        "type": "official-blog"
      }
    ],
    "publishedAt": "2026-05-16T00:00:00Z",
    "lastVerified": "2026-05-16",
    "methodologyVersion": "veritas-v0.1",
    "tags": [
      "constitutional-ai",
      "cai",
      "anthropic",
      "alignment",
      "harmlessness",
      "foundational",
      "2022",
      "introduced_in"
    ],
    "id": "6fa575eb9df5ac32",
    "statement": "Anthropic Constitutional AI Harmlessness introduced in paper: Bai et al. 2022 — training a helpful and harmless assistant."
  },
  "signature": {
    "algorithm": "HMAC-SHA256",
    "signedBy": "did:web:sourcescore.org",
    "signedAt": "2026-05-17T00:00:00.000Z",
    "signature": "e7c9817ddf4b019e642c34ab8e74a5e712d66e4964a0eb779a86899c30ca92ba"
  },
  "citedAs": "Anthropic Constitutional AI Harmlessness introduced in paper: Bai et al. 2022 — training a helpful and harmless assistant. — SourceScore Claim 6fa575eb9df5ac32 (verified 2026-05-16, signed e7c9817d…). https://sourcescore.org/claims/6fa575eb9df5ac32/"
}