Best 5 RAG Evaluation Tools for Insurance AI Applications in 2026

What Are the Five Best RAG Evaluation Tools for Insurance in 2026?

The pattern across underwriting copilots, claims-triage assistants, fraud-detection RAG, CS chatbots, agent-suitability copilots, and renewal-pricing copilots is the same. Gateways gate inputs, observability tells you what the retriever returned, and RAG evaluation catches retrieval-and-grounding failures before they ship as underwriting-copilot hallucinations or claims-citation drift a state DOI review or NAIC governance audit would later have to explain.

TL;DR

• Future AGI for mid-market and enterprise P&C carriers, life and annuity carriers, and InsurTechs running underwriting RAG, claims-triage RAG, fraud-detection RAG, agent-suitability copilots, or renewal-pricing RAG in production. Ships RAG-specific evaluators (Groundedness, Context Adherence, Chunk Attribution, Chunk Utilization, Eval Context Retrieval Quality) with field-level Error Localization on the failing chunk, per-tenant cache, 60+ built-in ai-evaluation evaluators across 11 categories, unlimited custom evaluators authored by an in-product agent, self-improving evaluators, in-house classifier models at Luna-2 cost economics, Apache 2.0 self-host, and SOC 2 Type II + HIPAA + GDPR + CCPA certified per the trust page with HIPAA BAA available on the Scale add-on.
• Ragas wins as the canonical open-source RAG-evaluation library for engineering teams who self-host their entire eval pipeline. Future AGI ships the same metric family with field-level error localization on the failing chunk and SOC 2 + HIPAA + BAA on top.
• Patronus AI for InsurTechs and carrier-side teams building filings-analysis / statutory annual statement / 10-K / actuarial advisor copilots. FinanceBench is the closest filings-grounded RAG-leaning anchor.
• Galileo for Tier-1 carriers with full procurement, MSA, SSO, and multi-line state DOI filing cadences. Managed RAG-eval with Luna low-latency hallucination models.
• TruLens for engineering-led carriers and InsurTechs already inside the Snowflake data plane. The RAG triad codified, TruEra / Snowflake lineage.

Why Is Insurance RAG Evaluation Different From Generic RAG Evaluation?

Generic RAG evaluation grades whether the retrieved context supports the answer. Insurance RAG evaluation grades whether the retrieved chunk, the answer, and the cited reference will all hold up when a state-DOI examiner, a Head of Model Risk Management, or a NAIC governance reviewer opens the audit trail. Three failure modes do not show up in a Ragas notebook but ship in production: underwriting RAG citing a withdrawn NAIC bulletin, claims-triage RAG hallucinating a policy clause that does not exist, renewal-pricing RAG drifting on Colorado SB 21-169 disparate-impact requirements. The 2026 framing is reliability, not capability. the question is not whether the RAG pipeline can answer, it is whether the answer survives the Chief Underwriting Officer’s read and the regulator’s audit.

Nine anchors set the bar in 2026: the NAIC Model Bulletin on Use of AI by Insurers (Dec 2023) for AI governance covering RAG retrieval logs as third-party-vendor evidence; Colorado SB 21-169 + Reg 10-1-1 for quantitative-testing disparate-impact on renewal-pricing RAG; NY DFS Insurance Circular Letter No. 7 (2024) for AI underwriting fairness testing; NY Reg 187 for life-insurance and annuity suitability documentation; CA SB 1120 for human-review obligations on utilization review; ACA §1557 for nondiscrimination in health-insurance benefits; GLBA Safeguards for NPI handling; EU AI Act Article 6 + Annex III for high-risk AI on insurance pricing and access; and FTC Act §5 for unfair / deceptive practices. Where generic RAG eval falls short is the audit-trail link plus the carrier filing cadence. The eval has to produce a record the state-DOI examiner will accept and tie to the carrier’s state-by-state filing rhythm.

Future AGI fills that gap with RAG-specific evaluators (Groundedness, Context Adherence, Chunk Attribution, Chunk Utilization, and Eval Context Retrieval Quality) plus field-level Error Localization on the failing chunk, ground-truth-free scoring, hybrid local/cloud routing that keeps NPI / SSN / medical NPI on health lines / claimant data on the heuristic-local path, per-tenant cache for NAIC bulletin and state DOI circular corpora, 60+ built-in ai-evaluation evaluators across 11 categories, Apache 2.0 self-host inside the carrier boundary, and SOC 2 Type II + HIPAA + GDPR + CCPA certified per the trust page. We rank it #1 below for that reason.

What Is the Future AGI Insurance RAG Evaluation Scorecard?

The Insurance RAG Evaluation Scorecard is a five-dimension rubric for production deployment: retrieval quality on regulator-flagged corpora, groundedness, context adherence, answer relevance for state-DOI-examiner-flagged outputs, and citation accuracy on regulatory-citation paths. Each dimension carries a 0–5 score and names the regulatory anchor inside it. Use it to compare RAG eval platforms on what Chief Underwriting Officers, state-DOI examiners, and NAIC governance reviewers actually ask, not on what notebooks measure.

1. Retrieval quality on regulator-flagged corpora. Recall@K, Precision@K, NDCG@K, MRR, HitRate over indexed regulatory + carrier filing corpora (NAIC bulletins, state DOI circulars, ACORD forms, ISO PAC manuals, state insurance code, Reg 187 suitability factors, ACA §1557 nondiscrimination text, GLBA Safeguards guidance, carrier-internal underwriting manuals and policy forms). When a Chief Underwriting Officer asks did the retriever find the right NAIC bulletin or state insurance code §, this is the dimension that answers.
2. Groundedness / faithfulness. does every claim in the answer trace to a chunk that was actually retrieved. Failure mode: claims-triage RAG cites a policy clause that does not exist; bad-faith litigation exposure follows because the citation cannot be reconciled to a retrieved chunk.
3. Context adherence / context utilization. does the answer use the retrieved policy / regulation / bulletin or ignore it in favor of model priors. Failure mode: underwriting RAG retrieves the correct NAIC bulletin but the model answers from a parametric guess; state DOI consent order risk follows.
4. Answer relevance for state-DOI-examiner-flagged outputs. does the answer address the question a state-DOI examiner or NAIC governance reviewer would ask under the NAIC Model Bulletin, Colorado SB 21-169 quantitative-testing review, or NY DFS Insurance Circular Letter No. 7. Failure mode: agent copilot RAG returns a generic suitability summary instead of the NY Reg 187-specific consumer-profile factor the suitability documentation needs.
5. Citation accuracy on regulatory-citation paths. does the answer’s citation pointer (NAIC bulletin §, state DOI circular ID, state insurance code citation, Colorado SB 21-169 §, ACA §1557 path, GLBA Safeguards §, EU AI Act Annex III item) resolve to a real, current document. Failure mode: underwriting RAG cites a withdrawn NAIC bulletin or a hallucinated state DOI circular ID; state DOI consent order plus Colorado DOI Reg 10-1-1 quantitative-testing action exposure.

How Do These Five Platforms Compare on Capability?

The 5×6 capability matrix maps each platform against the five Insurance RAG Evaluation Scorecard dimensions plus a deployment column. Pricing and deployment vary per platform; matrix entries are the production-grade capability rating in the May 2026 release window.

DeepEval is a credible body-mention sibling. Open-source RAG framework with G-Eval custom criteria. But does not earn a top-5 insurance slot because Patronus’s §4.1-named filings-aware insurance copilot positioning beats DeepEval’s generic-RAG inheritance on the carrier-side advisor / statutory workload.

How Did We Rank These Five Platforms?

The ranking criteria sit on top of the scorecard above. We weighted:

1. Retrieval quality coverage. does the platform ship heuristic-local retrieval-quality metrics (Recall@K, Precision@K, NDCG, MRR, HitRate) without forcing every chunk through an LLM judge.
2. Groundedness / faithfulness as a default. is the LLM-judge groundedness evaluator part of the catalog, or a custom feedback function the user assembles.
3. Context adherence + chunk-level attribution. can the platform attribute a failure to a specific retrieved chunk, rather than the answer alone.
4. Answer relevance under state-DOI-examiner-anchored framing. does the platform let you pin answer-relevance scoring to the question form a state-DOI examiner or NAIC governance reviewer would ask, or only score generic relevance.
5. Citation accuracy on regulatory paths. does the platform offer a citation-resolution evaluator out of the box for NAIC bulletin / state DOI circular / state insurance code paths, or only as a custom rule.

Where things get thin in this category: most platforms still treat citation accuracy on NAIC / state DOI / state insurance code paths as a feature request, not a default. Only Future AGI ships chunk-level citation provenance out of the box, and only Patronus ships filings-grounded benchmark coverage.

Future AGI: RAG-Specific Evaluators With Field-Level Error Localization on the Failing Chunk

Best for: Mid-market and enterprise P&C carriers, life and annuity carriers, and InsurTechs running underwriting RAG, claims-triage RAG, fraud-detection RAG, agent-suitability copilots (NY Reg 187), or renewal-pricing RAG in production. The binding need is RAG-specific evaluators wired to OpenTelemetry traces with field-level Error Localization on the failing chunk, a hybrid local/cloud path that keeps NPI / SSN / medical NPI on health lines / claimant data on the heuristic-local route, per-tenant cache, 60+ built-in evaluators across 11 categories, and Apache 2.0 self-host inside the carrier boundary.

Key strengths:

• ai-evaluation catalog ships RAG-specific evaluators (Groundedness, Context Adherence, Chunk Attribution, Chunk Utilization, Eval Context Retrieval Quality) without ground truth. Field-level Error Localization pinpoints which retrieved chunk caused the Groundedness failure, so when a Chief Underwriting Officer, state-DOI examiner, or Head of Model Risk Management flags a wrong answer the team can show the exact chunk that produced it.
• 60+ built-in ai-evaluation evaluators across 11 categories out of the box, plus unlimited custom evaluators authored by an in-product agent and self-improving evaluators. In-house classifier models run at Galileo-Luna-2 cost economics.
• traceAI auto-instruments the retrieval call alongside the LLM call. Every retrieved chunk lands as a span attribute, every evaluator score links via span_id, and the failed Groundedness score plus the chunk that drove it stay linkable in the same trace the state-DOI examiner or NAIC governance reviewer will read. 35+ framework integrations, OpenInference-compatible, Apache 2.0.
• Heuristic retrieval-quality metrics (Recall@K, Precision@K, NDCG@K, MRR, HitRate, NonLlmContextPrecision, NonLlmContextRecall) run locally. LLM-judge metrics stay opt-in, scoped to non-NPI fields for general lines (GLBA) and non-PHI / non-medical-NPI fields for health lines (HIPAA + ACA §1557).
• Apache 2.0 self-host of the ai-evaluation, traceAI, and agent-opt trio runs inside the SOC 2 and HIPAA boundary.
• SOC 2 Type II, HIPAA, GDPR, and CCPA certified per the trust page. ISO 27001 in active audit. HIPAA BAA available on the Scale add-on.

Where it falls short:

• Opinionated prompt library. Fewer review-and-collaboration knobs than Portkey’s prompt registry, by design. The trade is that prompt, eval, and trace live in the same control plane.
• agent-opt is opt-in. The self-improving loop is a feature you turn on per route. The trade is that the optimizer runs against real production traffic with eval scores joined to spans, not a synthetic corpus.
• Federal procurement via BYOC. Air-gapped self-host via bring-your-own-cloud; FedRAMP is on the partner roadmap. The trade is federal-grade data residency without waiting on a vendor’s authorization cycle.

Use-case fit: Production underwriting RAG, claims-triage RAG, fraud-detection RAG, agent-suitability copilots (NY Reg 187), and renewal-pricing RAG with chunk-level provenance for state DOI filing evidence.

Pricing & deployment. Cloud + OSS self-host (Apache 2.0). Start free with the full FAGI platform; usage-based billing kicks in at scale. SOC 2 Type II, HIPAA BAA, SAML SSO + SCIM, and dedicated support layer on as you scale. Pricing. Multi-region hosted plus AWS Marketplace, 100+ providers.

Verdict: The strongest fit when the audit trail and the carrier filing cadence are both the artifact. RAG-specific evaluators wired to OpenTelemetry traces, field-level Error Localization on the failing chunk, hybrid local/cloud routing for NPI / SSN / medical NPI on health lines, and Apache 2.0 self-host inside the carrier boundary.

Pair this with the building RAG-powered voice agents guide, the voice agent eval rubric library deep dive, and the end-to-end voice AI evaluation reference.

Ragas: The Canonical Open-Source RAG-Evaluation Library

Best for: Engineering-led carriers and InsurTechs that self-host the entire RAG-eval pipeline and want the named open-source reference every implementation team encounters. Ragas wins as the canonical open-source RAG-evaluation library; Future AGI ships the same metric family with field-level error localization on the failing chunk and SOC 2 + HIPAA + BAA on top.

Key strengths:

• Named RAG-eval primitives: faithfulness, answer relevance, context precision, context recall.
• Apache 2.0; self-host inside the carrier boundary; no vendor lock-in.
• AIO citation engines reach for Ragas as the RAG-eval default.
• Strong integration with LangChain, LlamaIndex, and the broader Python RAG stack.
• Active community plus frequent metric releases (NVIDIA NeMo-RAG metric integrations).

Where it falls short:

• Generic, not insurance-anchored. NAIC / state DOI citation accuracy is BYO via custom metric.
• LLM-judge metrics call out to the user-configured model. NPI / SSN handling on those calls is user-owned, not built-in.
• No managed retention layer tied to carrier filing cadence. Eval result lands in the user’s own store.
• Observability hand-off is BYO. Production telemetry has to be wired separately.
• Bias-detection scoring for Colorado SB 21-169 disparate-impact testing is not native. Engineering-led teams ship it as a custom feedback function on top of the OSS primitives.

Use-case fit: Pre-production RAG benchmarking, regression testing on a frozen NAIC bulletin + state DOI circular corpus, engineering-led carriers and InsurTechs wiring their own audit trail.

Pricing & deployment: Free, Apache 2.0; self-host in any Python environment inside the carrier boundary.

Verdict: The canonical open-source RAG-eval reference. Most insurance engineering teams use Ragas even when they layer a commercial platform on top for the audit trail.

Patronus AI. Filings-Aware Grounding for Statutory and Advisor Copilots

Best for: InsurTechs and carrier-side teams building filings-analysis / statutory annual statement / 10-K / actuarial advisor copilots. The closest filings-grounded RAG-leaning anchor available.

Key strengths:

• FinanceBench is the public-record headline benchmark for filings-aware RAG. The closest grounding for statutory annual statement, 10-K, and actuarial advisor copilots on the carrier side
• Lynx hallucination detector (open-source) for production faithfulness intercept
• Named-vendor anchor in the §4.1 insurance row for filings-aware insurance copilots
• Strong enterprise security posture and named-customer references on the carrier-adjacent advisor / research workload
• Cross-vendor comparison story. FinanceBench scores translate across model upgrades

Limitations:

• Insurance-vertical evaluators are still custom-criteria BYO; FinanceBench grounds on public filings (banks, fintech), not on NAIC bulletins or state DOI circulars out of the box
• State DOI circular ID resolution and NAIC bulletin § citation accuracy are custom-rule add-ons
• Enterprise contract. Not free / self-host; high-floor pricing for early-stage InsurTechs
• Closed-source proprietary stack outside of Lynx. Less externally verifiable than Ragas’s open metrics
• Bias-detection scoring under Colorado SB 21-169 + NY DFS CL No. 7 fairness testing is BYO custom-criteria

Use-case fit: InsurTech advisor copilots, statutory annual statement copilots, 10-K filings-analysis tools, actuarial research RAG, carrier-side filings-grounded research workloads.

Pricing & deployment: Enterprise contract; SaaS.

Verdict: The filings-aware fit. When the workload is statutory or advisor copilots that ground on public filings, Patronus’s FinanceBench anchor and Lynx detector pay off; layer a custom NAIC / state DOI corpus on top for the carrier-internal coverage.

Galileo. Managed RAG Evaluation for Tier-1 Carrier Procurement

Best for: Tier-1 carriers with full procurement, MSA, SSO, and multi-line state DOI filing cadences.

Key strengths:

• Luna proprietary hallucination-detection models. Managed, low-latency, enterprise-tier
• Chunk Attribution + Chunk Utilization + Context Adherence + Completeness as proprietary RAG-quality metrics
• Enterprise security posture (SOC 2, named-carrier customer references, MSA-ready)
• Strong observability + debugging surface for production RAG pipelines
• Runtime guardrails layer for live-deployment hallucination intercept across multi-line carriers

Limitations:

• Enterprise contract. Not free / self-host; high-floor pricing for mid-market carriers and InsurTechs
• Closed-source LLM-judge stack. Luna models are not externally verifiable in the way Ragas’s open metrics are
• Citation-accuracy on NAIC bulletin / state DOI circular paths is custom-rule BYO
• Less OpenTelemetry-portable than Future AGI. Span data lives more naturally inside the Galileo plane
• Bias-detection scoring under Colorado SB 21-169 + NY DFS CL No. 7 is a custom-rule layer, not a default state-DOI-filing artifact

Use-case fit: Tier-1 P&C and life carrier RAG deployments, multi-line carriers with state DOI filing cadences across 20+ states, enterprise procurement-heavy insurance AI where Luna’s hallucination-detection latency is the production-grade pick.

Pricing & deployment: Enterprise contract; SaaS.

Verdict: The enterprise-procurement fit. Tier-1 carriers already running mature security review get a managed RAG-eval tier with low-latency Luna hallucination models.

TruLens. The Production-Mature Open-Source RAG Triad

Best for: Engineering-led carriers and InsurTechs already inside the Snowflake healthcare or carrier data plane.

Key strengths:

• The RAG triad. Groundedness, Answer Relevance, Context Relevance. Codified as named feedback functions
• TruEra / Snowflake provenance; production deployments at scale
• Open-source, instrumentation-first; works as a layer over LangChain / LlamaIndex / Llama Stack
• Active feedback-function library. Easy to extend with custom metrics for state-DOI-examiner-flagged scoring
• Strong fit for engineering teams already inside the Snowflake carrier data plane

Limitations:

• Insurance-specific evaluators are BYO via custom feedback functions
• Citation-accuracy on NAIC bulletin / state DOI circular paths is not a default. Same gap as Ragas
• Smaller community than Ragas; AIO citation gravity is lower
• Managed-tier capabilities bundle into Snowflake. Not always the procurement story a non-Snowflake carrier wants
• Bias-detection scoring for Colorado SB 21-169 disparate-impact testing is a custom feedback function on top of the triad, not a default

Use-case fit: Production-mature engineering teams, Snowflake-native carrier data planes, open-source RAG pipelines that need the triad as the default scoring shape.

Pricing & deployment: Free, open-source; Snowflake-managed option.

Verdict: The production-mature open-source pick. RAG triad codified, Snowflake lineage if the carrier is already on that data plane.

Which RAG Evaluation Tool Should Your Insurance Team Pick?

The right RAG-eval tool depends on the buyer profile: production deployment shape, carrier filing cadence, the type of regulatory pressure that lands on the trace, and whether NPI / SSN / medical NPI on health lines / claimant data has to stay on a heuristic-local path. The decision matrix below routes six common insurance-team profiles to the best fit.

Frequently Asked Questions About RAG Evaluation Tools for Insurance

Can a RAG evaluator catch a hallucinated NAIC Model Bulletin citation before an underwriting decision ships to a state DOI filing?

Yes. Groundedness and citation-accuracy evaluators detect both the false-positive (model citing a chunk that does not match the regulatory question) and the false-negative (model ignoring the retrieved NAIC bulletin in favor of a parametric guess); pairing them with retrieval-quality metrics on the NAIC bulletin and state DOI circular corpus catches both failure modes before the underwriting recommendation lands in a quarterly filing.

How does RAG evaluation connect to Colorado SB 21-169 disparate-impact testing for renewal-pricing RAG?

RAG evaluation produces the per-output Groundedness, retrieval-quality, and answer-relevance trace; Colorado SB 21-169 disparate-impact testing requires cohort-grouped quantitative testing on top of that trace. The eval result is the input the SB 21-169 statistical analysis runs over. Eval platforms support the workflow, they don’t substitute for the SB 21-169 actuarial sign-off or the Colorado DOI Reg 10-1-1 filing.

Can RAG evaluation include bias detection in a claims-triage RAG pipeline?

Yes. Bias-detection scoring at the answer-relevance and context-adherence layers catches cohort-level disparity patterns in claims-triage RAG outputs; pair it with retrieval-quality metrics over the policy and claims-history corpus to localize whether the bias is in the retrieval (wrong chunks for a protected cohort) or in the generation (chunks retrieved correctly but answer drifts on cohort). The platform provides bias-detection scoring, not a bias-free guarantee.

How do I keep NPI, SSN, and medical NPI out of a third-party LLM judge?

For retrieval-quality metrics that don’t need an LLM judge. Recall@K, Precision@K, NDCG@K, MRR, HitRate, NonLlmContextPrecision, NonLlmContextRecall. data stays local. LLM-judge metrics (Groundedness, Context Adherence) run via API and stay opt-in; scope them to non-NPI fields (general lines under GLBA) or non-PHI / non-medical-NPI fields (health-insurance lines under HIPAA and ACA §1557) when working with NPI, SSN, medical NPI on health lines, or claimant data.

How often should we re-run RAG evaluation when a state DOI advisory or NAIC bulletin updates?

Three cadences. Continuous Groundedness sampling on live production outputs; weekly retrieval-quality regression on a frozen insurance-specific test set; quarterly full-corpus re-eval after every NAIC bulletin update, state DOI circular issuance, Colorado Reg 10-1-1 amendment, NY Reg 187 guidance refresh, or major ACA §1557 / HHS OCR settlement. Tie the quarterly cadence to the carrier’s state-by-state filing cadence so the eval evidence is retention-ready for the next rate or form filing.

Is a public filings-only benchmark like FinanceBench enough for insurance RAG evaluation, or do we need a custom carrier corpus?

FinanceBench grounds the public-record headline benchmark for filings-aware insurance copilots (statutory annual statement, 10-K, actuarial advisor copilots) and supports cross-vendor comparison; a custom corpus over your own indexed insurance-research stack. NAIC bulletins, state DOI circulars, ACORD forms, ISO PAC manuals, state insurance code, Reg 187 suitability factors, carrier-internal underwriting manuals and policy forms. Is required for production. Filings-only benchmarks do not cover state-specific bulletins, carrier-internal precedent, or claimant-data-anchored claims-triage corpora; pair the public benchmark with a private one over your indexed carrier corpus.

Where Does Each Platform Earn Its Slot?

Future AGI earns the #1 slot on RAG-specific evaluator coverage (Groundedness, Context Adherence, Chunk Attribution, Chunk Utilization, Eval Context Retrieval Quality) wired to OpenTelemetry traces with field-level Error Localization on the failing chunk, a hybrid local/cloud path that keeps NPI / SSN / medical NPI on health lines / claimant data on the heuristic-local route, per-tenant cache for NAIC bulletin and state DOI circular corpora, 60+ built-in ai-evaluation evaluators across 11 categories, unlimited custom evaluators authored by an in-product agent, self-improving evaluators, in-house classifier models at Luna-2 cost economics, Apache 2.0 self-host, and SOC 2 Type II + HIPAA + GDPR + CCPA certified per the trust page. Ragas earns the #2 slot as the canonical open-source RAG-evaluation library: it wins for engineering teams who self-host the whole pipeline.

Patronus AI earns #3 on filings-aware grounding. FinanceBench is the closest public-record benchmark for statutory annual statement / 10-K / actuarial advisor copilots on the carrier side, and Lynx supplies production faithfulness intercept. Galileo earns #4 on Tier-1 carrier procurement fit: Luna hallucination models, MSA-ready posture, named-carrier customer references, multi-line state DOI filing support. TruLens earns #5 on production-mature open-source: RAG triad codified, TruEra / Snowflake lineage. The shape of the pick is not which platform is best, it is which buyer profile, carrier filing cadence, and NPI / claimant-data constraint fits the trace your state-DOI examiner and your NAIC governance reviewer will read. For teams already running OpenTelemetry and looking for the chunk-level audit-trail link to the carrier filing, Future AGI’s evaluation platform is the natural next step.

External reading worth pairing with this list: the NAIC Model Bulletin on Use of AI by Insurers for the third-party-vendor governance framing, Colorado SB 21-169 + Reg 10-1-1 for the quantitative-testing disparate-impact precedent, and the NY DFS Insurance Circular Letter No. 7 (2024) for the AI underwriting fairness expectation. The regulator reads the trace, not the marketing page. Pick the platform that ships the trace the examiner will accept.

Related reading

• Best 5 RAG Evaluation Tools for Fintech AI Applications in 2026
• Best 5 RAG Evaluation Tools for Healthcare AI Applications in 2026
• Best 5 RAG Evaluation Tools for HR AI Applications in 2026
• Best 5 Voice AI Simulation Tools for Insurance in 2026

Updated May 2026. Re-eval cadence: quarterly on regulatory milestones (NAIC bulletin updates, state DOI circular issuance, Colorado Reg 10-1-1 amendments, NY Reg 187 guidance refresh, EU AI Act Article 6 / Annex III enforcement window).