About & Sources

🧠 Our Research Methodology

Every claim on this site is vetted through a five-layer process. We don't trust any single AI model, any single source, or any single reviewer — the layers are designed to catch what each other miss.

Layer 1 — Mixture-of-Experts (MoE) draft: Grok (xAI) with live web and X search, Gemini 2.5 Pro (Google) with Google Search grounding, and Claude Opus (Anthropic) with web search each produce an independent draft. DeepSeek-V3.2 reviews for logical consistency, date accuracy, and quantitative errors. Cerebras Llama synthesizes the cross-model consensus.

Layer 2 — Primary-source verification: Every factual claim is cross-checked against authoritative primary sources (see the full list below — e.g., CDC, NIH, ACOG, NAIC, CFPB, EPA, FDA, peer-reviewed literature). Claims without a dated primary source are either re-researched or cut.

Layer 3 — Publication gate: A claim ships only when all five experts AGREE and at least one primary source substantiates it. Disagreements trigger another round of re-synthesis. Unresolved disputes are flagged, not hidden.

Layer 4 — Counterfactual review: Every surviving claim is then re-tested against its strongest opposing evidence — we actively search for the counter-argument or dissenting primary source that would falsify the claim. Where the contrary evidence is substantial, the claim is labeled contested or emerging rather than presented as settled.

Layer 5 — Firm review: The consolidated output then passes through our own editorial understanding. We ask whether each claim is genuinely useful to a reader (not just technically correct), rewrite for clarity, and surface the practical implication. We're technologists, not clinicians or attorneys — what we contribute at this layer is reader utility, not medical or legal authority.

📚 Sources & Citations

Content on this site is drawn from and cross-checked against the following authoritative sources:

1. National Association of Insurance Commissioners (NAIC)
2. Insurance Information Institute (III)
3. Consumer Federation of America — Insurance
4. U.S. Department of the Treasury — Federal Insurance Office
5. State Insurance Departments (all 50 + DC)
6. AM Best financial strength ratings
7. S&P Global Ratings — Insurance
8. Moody's Insurance ratings
9. NCCI (National Council on Compensation Insurance)
10. ISO / Verisk actuarial data
11. CMS — Medicare & Medicaid
12. healthcare.gov
13. International Association of Insurance Supervisors (IAIS)
14. Lloyd's of London market reports
15. Swiss Re and Munich Re sigma / NatCat reports

Where sources disagree, we surface the disagreement rather than pick a side. Where evidence is weak or emerging, claims are labeled accordingly.

✍️ Editorial Policy

We update pages whenever new primary research emerges or a reader flags an error. Each page is dated last reviewed. Corrections welcomed: [email protected].

We do not accept paid placement or sponsored content in the main knowledge base.

🏛️ About the Publisher

Published by AI Pulled — a New York firm running a portfolio of MoE-built consumer knowledge bases across health, home, finance, legal, and parenting topics. We're technologists, not clinicians, attorneys, or licensed financial advisors: our content is educational and carries that disclosure on every page.

Our editorial principle is content quality over content volume. We update claims as the evidence changes, retire pages that no longer meet our bar, and treat reader corrections as a primary input to the review cycle.

Corrections, counter-evidence, and source challenges are welcome and read: [email protected] · aipulled.com