Pet Insurance Ratings: What AI Changes for Underwriting

AM Best doesn’t move a company’s rating status for fun. So when it removes ratings “from under review” and affirms an A- (Excellent), that’s a signal the insurer has answered hard questions about stability, capital, and risk controls.

That’s exactly what happened this week with Independence American Insurance Co. (IAIC), a major pet health insurer operating under Independence Pet Holdings. AM Best affirmed IAIC’s Financial Strength Rating of A- and Long-Term Issuer Credit Ratings of “a-”, with a stable outlook, after the company re-implemented quota share reinsurance and received an additional $125 million capital contribution from its parent in Q3 2025.

For an “AI in Insurance” series, this is more than a ratings update. It’s a real-time case study in what trust looks like in an emerging line like pet insurance—and why AI underwriting, AI-powered risk assessment, and analytics-driven transparency are becoming table stakes as carriers scale.

What AM Best’s decision really says about pet insurance risk

AM Best’s decision boils down to this: the risk story became clearer and more controllable. Ratings agencies want to see that a fast-growing insurer can keep its promises under stress—especially in product lines where historical loss patterns are still settling.

In IAIC’s case, AM Best pointed to several concrete drivers:

• Reinsurance structure normalized: IAIC renegotiated and re-implemented its pet insurance quota share contracts in Q3 2025.
• Capital support was explicit: Independence Pet Holdings contributed $125 million.
• Balance sheet strength is “very strong”: AM Best cited liquidity, an investment portfolio it views favorably, and no debt in the financial structure.
• Profitability showed up in the numbers: IAIC reported $44.3 million net income through Q3 2025, driven by underwriting results and investment gains.

There’s a broader message for pet insurers and insurtechs: growth doesn’t impress ratings agencies unless your risk and capital story scales with it.

Why pet insurance is uniquely “ratings sensitive”

Pet insurance is still a high-growth, relatively young market compared to auto or homeowners. That creates three recurring challenges:

1. Thin credibility in long-term loss patterns (especially for newer product designs and rapidly changing vet pricing)
2. Acquisition-driven growth that complicates data consistency (IAIC has completed multiple pet-focused acquisitions)
3. Reinsurance dependence to manage volatility while scaling

Ratings firms notice all three—and so do sophisticated distribution partners.

The hidden mechanics: reinsurance, BCAR, and why transparency wins deals

If you only skim ratings announcements, it’s easy to miss the most important line: AM Best said IAIC’s risk-adjusted capitalization, measured by Best’s Capital Adequacy Ratio (BCAR), improved “to the strongest level” after the reinsurance re-implementation and the capital contribution.

Here’s the practical interpretation: BCAR is a scoreboard for whether capital is adequate for the insurer’s risk profile. When BCAR weakens, it’s usually not because a company is “bad.” It’s because the relationship between:

• net retained risk,
• reinsurance credit / structure,
• premium growth,
• reserves,
• investment risk,
• and operational risk

…has drifted into a less comfortable zone.

AM Best also noted that IAIC’s BCAR decline at year-end 2024 related to the refiling of annual statements to reflect a deposit account on reinsurance contracts effective Jan. 1, 2024. Translation: in fast-moving reinsurance programs, even accounting treatment and structure can materially change the way capital adequacy is viewed.

What AI changes: explaining capital and reinsurance in plain English

Most carriers underestimate this: “Transparency” isn’t a PDF report. It’s the ability to answer follow-up questions fast, with numbers that reconcile.

AI can help in three practical ways:

1. Automated reconciliations and anomaly detection

   • Flag mismatches across policy admin, claims, finance, and reinsurance bordereaux.
   • Catch “data drift” after acquisitions when field definitions don’t align.

2. Scenario modeling that’s actually usable

   • Run stress tests on retention changes (e.g., quota share shifts effective mid-year like IAIC’s July 1, 2025 change).
   • Quantify impacts to capital needs, earnings volatility, and liquidity.

3. Explainable reporting for non-technical stakeholders

   • Generate management-ready narratives: what changed, why it changed, and what the insurer did about it.

If you want a snippet-worthy truth: ratings stability is often a data problem before it becomes a capital problem.

AI underwriting for pet insurance: where it helps (and where it can backfire)

AI underwriting in pet insurance works best when it does one job extremely well: turn messy, incomplete signals into consistently priced risk decisions.

AM Best highlighted IAIC’s scale and growth—premium growth over five years, acquisitions (including Pets Best, Felix, and a major pet portfolio), and a mix of individual and group coverage. Growth like that increases operational complexity. AI can reduce that complexity, but only if it’s built around insurance reality.

High-impact AI use cases for pet insurers

1) Pricing and risk selection with better feature engineering

Pet insurance loss costs are influenced by inputs that aren’t always captured cleanly:

• breed and breed-mix proxies
• geography and local veterinary cost inflation
• age and pre-existing condition indicators
• benefit design (deductibles, reimbursement %, annual limits)
• renewal behavior (anti-selection risk)

A solid machine learning pricing model can incorporate these factors more consistently than manual rating adjustments—especially when paired with strong governance.

2) Claims triage and clinical-text intelligence

A huge portion of pet claims value is trapped in unstructured vet notes, invoices, and medical records. Natural language processing can:

• extract diagnoses and procedures
• detect missing documentation early
• reduce cycle time and leakage
• standardize coding across provider styles

This matters because pet insurers often win or lose on service experience. Claim speed is a retention lever.

3) Renewal and retention models that reduce volatility

As portfolios mature, profitability is as much about retaining the right risks as it is about acquiring new ones. AI can help identify:

• customers likely to lapse due to price sensitivity
• policies at risk of adverse selection
• cohorts where plan redesign or communications reduce churn

Where AI can backfire: three failure modes I see often

1. “Black box” pricing without audit trails

   • If you can’t explain a rate change internally, you won’t explain it to regulators, partners, or rating analysts.

2. Training data that doesn’t survive acquisitions

   • IAIC’s acquisition history is common in pet insurance. Models trained on one portfolio can misprice another unless you normalize data and validate performance by cohort.

3. Automation that ignores reinsurance structure

   • Underwriting decisions change the shape of retained risk. If pricing teams don’t coordinate with reinsurance and capital management, you get unpleasant BCAR surprises.

A stance worth taking: AI doesn’t replace underwriting discipline—it punishes you for not having it.

Trust is the product: why ratings, partners, and AI governance are linked

IAIC operates with multiple brands and distribution relationships, including white-label and co-branded programs. In pet insurance, that distribution model is common—and fragile.

Here’s why: partners don’t just care about loss ratios. They care about reputational and continuity risk:

• Will claims get paid quickly in a surge?
• Will pricing remain stable at renewal, or will customers see sharp increases?
• Will the insurer stay financially strong enough to support long-duration policies?

AM Best calling out enterprise risk management as “appropriate” is a subtle but meaningful endorsement. It suggests processes exist to identify and control risk, not just react to it.

What “AI transparency” looks like in practice

If you’re building or buying AI systems for underwriting or claims, transparency should be designed in from day one:

• Model cards that document training data, target variable, limitations, and approval owners
• Drift monitoring tied to actionable thresholds (not dashboards nobody checks)
• Fairness and compliance testing appropriate to your jurisdiction and product
• Human override workflows so underwriters can intervene and the system learns responsibly
• Reinsurance-aware reporting that shows net retained risk by segment, not just gross growth

This is how AI supports ratings outcomes: it makes risk decisions legible—to executives, auditors, regulators, and rating analysts.

A practical checklist for pet insurers preparing for scrutiny

If a ratings review, reinsurance renewal, or capital discussion is on your 2026 calendar, use this checklist to tighten your story.

Data and analytics readiness

• Do you have a single definition of “net premium” across finance, actuarial, and reinsurance?
• Can you reconcile claims counts and paid loss between systems within days, not weeks?
• After acquisitions, have you normalized fields (breed, age, diagnosis coding) and tracked residual gaps?

Underwriting and pricing controls

• Can you explain your top 10 pricing drivers in plain language?
• Are rate changes tested for portfolio impact and operational impact (call center load, complaint risk)?
• Do you have controls to prevent silent shifts in appetite due to model updates?

Capital and reinsurance alignment

• For every distribution channel, do you know your retained loss volatility under your current quota share?
• Do you run stress tests for veterinary cost inflation and claims severity spikes?
• Can you show how underwriting actions change capital needs over the next 12–24 months?

If you can answer these cleanly, you’re not just “AI-enabled.” You’re ratings-ready.

Where this is heading in 2026: AI as the credibility layer

Pet insurance is consolidating, scaling, and getting more professional fast. The IAIC update shows how the market is maturing: ratings stability is being earned through capital actions, reinsurance structure, and operational performance—not hype.

For the broader “AI in Insurance” narrative, this is the next phase. AI isn’t a flashy add-on; it’s becoming the credibility layer that supports:

• more consistent underwriting decisions,
• clearer risk reporting,
• better claims experiences,
• and faster, more defensible responses when outside stakeholders ask tough questions.

If you’re a pet insurer (or a carrier entering pet) and you want growth that doesn’t trigger constant scrutiny, focus on one outcome: make your risk story measurable and explainable.

Where are you most exposed right now—pricing accuracy, claims leakage, reinsurance reporting, or acquisition data integration?