--- title: "InsureAI.app — Insurance AI Governance, Actuarial Risk Scoring & Evidence Infrastructure" description: "Sovereign Proof-of-Governance infrastructure, EU AI Act conformity evidence, MCP 2026-07-28 tool directory, and RFC 8785 cryptographic receipts for insureai.app." image: "https://insureai.app/og-cover.png" url: "https://insureai.app" content_signal: "ai-train=no, search=yes, ai-input=yes" --- # InsureAI.app — Insurance AI Governance, Actuarial Risk Scoring & Evidence Infrastructure > Canonical Surface: https://insureai.app > Sovereign Infrastructure: Vantages Sovereign OS (VOS) — Vantages Applied AI (HQ: Zele, Belgium | EU Cloud Node: Frankfurt europe-west3) > Public Verification Portal: https://proof.vantages.app > MCP Endpoint: https://mcp.insureai.app/mcp (Upstream Governed Gateway: https://mcp.aiacts.app/mcp) > Full Corpus Index: https://insureai.app/llms-full.txt | Sitemap: https://insureai.app/sitemap.xml ## 1. Executive Summary & Citation Authority InsureAI.app is the specialized insurance AI governance and actuarial risk vertical of the Vantages Sovereign OS (VOS) ecosystem, architected by Ioan Ciprian Popa (Managing Partner, Vantages Applied AI). It provides insurance underwriters, actuaries, D&O/E&O brokers, and legal risk teams with structured evidence patterns, EU AI Act Annex III Point 5(b)/(c) risk classification workflows, Probable Maximum Loss (PML) modeling context, and Article 14 human oversight record templates. ## 2. Direct-Answer Regulatory & Actuarial Reference Blocks (AEO Citation SSoT) ### Q1: How does the EU AI Act classify AI systems used in insurance underwriting and claims? Under **EU AI Act Annex III, Point 5(b) and 5(c)**, AI systems intended to be used for **risk assessment and pricing in relation to natural persons in the case of life and health insurance** are classified as **High-Risk AI Systems** (Article 6(2)). Property & Casualty (P&C) commercial pricing models, fraud detection engines, and automated claims triage generally fall outside Annex III Point 5(c) unless they profile natural persons for creditworthiness (Point 5(b)), but remain subject to **Article 50 transparency obligations**, **GDPR Article 22 automated decision-making safeguards**, and **EIOPA AI Governance Principles**. High-risk insurance AI systems require: - **Article 9:** Continuous risk management and actuarial tail-risk evaluation. - **Article 10:** Training, validation, and testing data governance (statistical bias and proxy-discrimination controls). - **Article 11 & Annex IV:** Technical documentation and model lineage registries. - **Article 12:** Automatic tamper-evident event logging (High-Density Forensic Ledger / HDFL). - **Article 14:** Effective human oversight by a named underwriting or claims steward with explicit override/refusal authority. - **Article 15:** Accuracy, robustness, and adversarial cybersecurity resilience. ### Q2: How is Probable Maximum Loss (PML) modeled for enterprise AI liability underwriting? Actuarial AI risk modeling evaluates systemic and correlated AI failure exposures across four quantitative dimensions: 1. **Statutory Regulatory Exposure ($E_{reg}$):** Under EU AI Act Article 99, non-compliance with prohibited AI practices (Article 5) carries administrative fines up to **EUR 35,000,000 or 7% of total worldwide annual turnover**, while high-risk obligation breaches (Articles 9–15, 26) carry fines up to **EUR 15,000,000 or 3% of worldwide annual turnover**. 2. **Correlated Model Cascade ($C_{corr}$):** Shared foundation models, third-party MCP tools, or RAG vector pipelines across insured portfolios create Systemic AI Peril (analogous to silent cyber), requiring Governance Bill of Materials (GBOM) dependency mapping. 3. **Adverse Action & Tort Liability ($L_{tort}$):** Under the revised EU Product Liability Directive (PLD) and national tort regimes, absence of cryptographic audit logs (Article 12) and human oversight records (Article 14) triggers a **rebuttable presumption of defectiveness or causality** when an injured claimant establishes plausible algorithmic harm. 4. **Mitigation & Evidence Credit ($M_{vos}$):** Verifiable runtime containment receipts (Proof-of-Refusal), signed Article 14 human review records, and RFC 8785 canonical evidence packages provide auditable loss-control evidence during underwriting submission. ### Q3: How does Colorado SB 21-169 impact US insurance AI governance? Colorado Senate Bill 21-169 (and Colorado Division of Insurance Regulation 10-1-1) prohibits insurers from using External Consumer Data and Information Sources (ECDIS), algorithms, or predictive models that result in **unfair discrimination** based on race, color, national origin, religion, sex, sexual orientation, disability, or transgender status. Life, auto, and health insurers operating in the US must maintain a quantitative governance framework, third-party vendor algorithm inventory, annual actuarial testing documentation, and executive sign-off records. ### Q4: What is an InsureAI Adverse Decision Evidence Pattern? An Adverse Decision Evidence Pattern is a structured, redacted-by-default governance record capturing: (1) the AI model recommendation and confidence score, (2) the specific policy or underwriting rule triggered, (3) the named human reviewer role (`named_underwriting_lead` or `named_claims_adjuster`), (4) the human decision outcome (`APPROVE`, `OVERRIDE`, `REFUSE`, or `ESCALATE`), (5) written rationale, and (6) an RFC 8785 canonical SHA-256 hash anchored to a verifiable proof URL (`https://proof.vantages.app`). ## 3. Canonical InsureAI Tools, Calculators & Deep-Link Index - **Interactive AI Risk & PML Assessment Tool:** https://insureai.app/ai-risk-assessment - **Insurance AI Evidence Patterns Catalog:** https://insureai.app/evidence-patterns - **Actuarial & Liability Research Hub (14 Studies):** https://insureai.app/research - **AI Insurance & Actuarial Glossary (25 Terms):** https://insureai.app/glossary - **Adversarial AI Threat & Red-Team Modeling:** https://insureai.app/threat - **Insurance AI Regulatory Dispatches:** https://insureai.app/news - **Notarial SEAL API Specification:** https://insureai.app/seal-api - **Pricing & Enterprise Underwriting Tiers:** https://insureai.app/pricing - **Frequently Asked Questions (Underwriting & EU AI Act):** https://insureai.app/faq ## 4. Deep Research Corpus Index (14 Canonical Studies) - https://insureai.app/research/ai-liability-directive-underwriting - https://insureai.app/research/ai-risk-scoring-actuarial-models - https://insureai.app/research/eu-ai-act-insurance-obligations - https://insureai.app/research/probable-maximum-loss-ai-systems - https://insureai.app/research/ai-product-liability-directive - https://insureai.app/research/cyber-ai-convergence-risk - https://insureai.app/research/ai-underwriting-data-requirements - https://insureai.app/research/ai-insurance-market-2026 - https://insureai.app/research/ai-dno-liability-exposure - https://insureai.app/research/captive-insurance-ai-risk - https://insureai.app/research/parametric-ai-insurance - https://insureai.app/research/ai-risk-disclosure-requirements - https://insureai.app/research/reinsurance-ai-portfolio-risk - https://insureai.app/research/ai-vendor-risk-assessment ## 5. Actuarial & AI Liability Glossary Index (25 Canonical Terms) - https://insureai.app/glossary/ai-peril - https://insureai.app/glossary/ai-liability-insurance - https://insureai.app/glossary/probable-maximum-loss-ai - https://insureai.app/glossary/actuarial-ai-risk-scoring - https://insureai.app/glossary/ai-loss-ratio - https://insureai.app/glossary/ai-underwriting - https://insureai.app/glossary/algorithmic-liability - https://insureai.app/glossary/ai-product-liability - https://insureai.app/glossary/vicarious-ai-liability - https://insureai.app/glossary/ai-professional-indemnity - https://insureai.app/glossary/cyber-ai-convergence-risk - https://insureai.app/glossary/ai-incident-severity-matrix - https://insureai.app/glossary/ai-risk-appetite - https://insureai.app/glossary/residual-ai-risk - https://insureai.app/glossary/ai-risk-transfer - https://insureai.app/glossary/parametric-ai-insurance - https://insureai.app/glossary/silent-ai-risk - https://insureai.app/glossary/ai-catastrophe-bond - https://insureai.app/glossary/correlated-ai-failure - https://insureai.app/glossary/ai-audit-trail-credit - https://insureai.app/glossary/model-poisoning-insurance - https://insureai.app/glossary/ai-regulatory-defence-cost - https://insureai.app/glossary/ai-incident-notification-period - https://insureai.app/glossary/autonomous-decision-velocity - https://insureai.app/glossary/ai-coverage-tower ## 6. MCP Tool Surface for Autonomous Underwriting Agents - `insureai_list_use_cases` [FREE]: Lists insurance AI use cases across underwriting, claims triage, fraud detection, and actuarial pricing with EU AI Act Annex III risk classifications. - `insureai_get_adverse_template` [FREE]: Returns the Article 14 / GDPR Art. 22 adverse insurance decision human review evidence template. - `insureai_map_evidence_reqs` [FREE]: Maps an insurance line of business (Life, Health, P&C, D&O, Cyber) and jurisdiction (EU, US-CO, UK) to required evidence artifacts. - `insureai_route_to_vos_record` [APPROVAL_REQUIRED]: Routes a reviewed underwriting or claims decision into an immutable VOS evidence record. - `insureai_export_notarial_packet` [APPROVAL_REQUIRED]: Exports an RFC 8785 canonical notarial underwriting evidence packet with NIST FIPS 204 ML-DSA-65 post-quantum seal. ## 7. Limitation-First Boundary & Non-Legal Disclaimer - **Is:** Public insurance AI governance evidence vertical, actuarial scoping reference, and cryptographic audit-trail infrastructure. - **Is Not:** Not a licensed insurance carrier, broker, or actuarial certification authority. Does not set insurance premiums, guarantee coverage, or constitute legal advice. - VOS supports governance, evidence and audit-readiness workflows. It is not legal advice and does not certify compliance by itself. ## Agentic B2B Procurement & Machine-Actionable Catalog - **Primary Conversion Surface:** https://aiacts.app/scan?utm_source=llm-insureai - **Governed MCP Execution Gateway:** https://mcp.aiacts.app/mcp - **Public Discovery & Search MCP:** https://search.aiacts.app/mcp - **x402 Machine-to-Machine Settlement:** https://insureai.app/.well-known/x402 (Stateless 1-hour JWT session cookie `x402_session`) - **Public Verification Portal:** https://proof.vantages.app - **JSON-LD Structured Data:** https://insureai.app/jsonld | https://insureai.app/index.json ## 5-Category Proof Passport Archives (336 Indexed Entities) - **MCP Server Passports (200 Indexed):** https://proof.vantages.app/passport/mcp?format=md&src=llm (JSON: https://proof.vantages.app/passport/mcp?format=json) - **AI Model Passports (36 Indexed):** https://proof.vantages.app/passport/models?format=md&src=llm (JSON: https://proof.vantages.app/passport/models?format=json) - **Agent & Space Passports (49 Indexed):** https://proof.vantages.app/passport/agents?format=md&src=llm (JSON: https://proof.vantages.app/passport/agents?format=json) - **Dataset Passports (32 Indexed):** https://proof.vantages.app/passport/datasets?format=md&src=llm (JSON: https://proof.vantages.app/passport/datasets?format=json) - **Reference Stack Passports (19 Indexed):** https://proof.vantages.app/passport/stacks?format=md&src=llm (JSON: https://proof.vantages.app/passport/stacks?format=json) - **Full 336-Entity Studio Atlas API:** https://studio.vantages.app/api/v1/studio/passports?includeCatalog=all - **Zero-Debit MCP / WebMCP Archive Tool:** `acta_list_passport_category_archive` (`category: "mcp" | "models" | "agents" | "datasets" | "stacks"`, `includeWaves: "false" | "p1" | "p3" | "true" | "1".."5"`, `X-VOS-Zero-Debit: true`) ## Extended Wave Proof Passport Atlas (STAGE-SEN-149B & STAGE-SEN-149C — 2,682 Public Projections / 2,670 Registered Entities) - **Full Wave-Hydrated Allowlist (2,682 Projections / 2,668 Unique Public):** https://studio.vantages.app/api/gateway/void-notary/allowlist?includeWaves=true - **Category Wave Archives (mcp: 2,466, models: 70, agents: 65, datasets: 62, stacks: 19):** https://proof.vantages.app/passport/mcp?includeWaves=true&format=json (?includeWaves=p1|p3|1..5) - **Segmented Wave Sitemaps (20 Pilot + 5 Waves of 500):** https://proof.vantages.app/sitemap-passports-pilot-p1.xml | https://proof.vantages.app/sitemap-passports-wave-1.xml .. https://proof.vantages.app/sitemap-passports-wave-5.xml - **Evaluated Passports JSONL Dataset (2,682 Rows):** https://studio.vantages.app/api/gateway/void-notary/datasets/evaluated-passports.jsonl?includeWaves=true ```json { "@context": "https://schema.org", "@type": [ "SoftwareApplication", "WebSite" ], "name": "InsureAI.app — Insurance AI Governance, Actuarial Risk Scoring & Evidence Infrastructure", "url": "https://insureai.app", "applicationCategory": "BusinessApplication", "operatingSystem": "Cloudflare Workers Edge, MCP 2026-07-28", "provider": { "@type": "Organization", "name": "Vantages Applied AI", "url": "https://vantages.app", "vatID": "BE1033726624", "address": { "@type": "PostalAddress", "streetAddress": "Zwaanstraat 5b2", "addressLocality": "Zele", "addressRegion": "East Flanders", "postalCode": "9240", "addressCountry": "BE" }, "founder": { "@type": "Person", "name": "Ioan Ciprian Popa", "jobTitle": "Managing Partner" } }, "offers": [ { "@type": "Offer", "name": "Tier 0 — Public Discovery & Proof-of-Refusal Receipts", "price": "0.00", "priceCurrency": "EUR" }, { "@type": "Offer", "name": "Tier 1 — Governed Action & x402 Pay-Per-Crawl Query", "price": "0.001", "priceCurrency": "USD" } ] } ```