The actuarial basis for AI underwriting.
AI E&O. Cyber-AI extensions. Model-liability endorsements. The category lacks 200 years of claims tables — but it doesn't have to fly blind. We provide structural-audit infrastructure that turns AI risk into something a carrier can actually price.
The category problem
AI liability is uninsurable when it's unpriceable.
Traditional insurance had two centuries of claims experience before AI showed up. Your underwriters have weeks. Without an actuarial substrate, AI policies either get declined, priced punitively, or hollowed out with exclusions until the coverage is meaningless.
What we provide
Six capabilities, one infrastructure
Quote with evidence, not guesswork
Run a structural audit on the model before issuing the policy. Verdict, blind-spot map, and coverage analysis become pricing-grade input for your underwriter. SAFE → standard rate; RED_FLAG → loaded or decline; BAD_MODEL → fix and reapply.
Risk-tier the AI policies you already write
Extend your existing AI E&O, Cyber, or general-liability product with structural-audit-based tiering. Lower premiums for clean models, higher for fragile ones — defensible to brokers, transparent to insureds, auditable to reinsurers.
Catch drift before the loss
Periodic re-audits during the policy term. Structural drift past threshold triggers automatic premium adjustment, mid-term cancellation rights, or non-renewal. The same evidence stream feeds both your portfolio team and your reinsurer reporting.
Forensic record when a loss occurs
Every audit is timestamped and signed. When a claim is filed, the audit history shows whether the model was within policy parameters at the time of the incident — documented evidence for adjudication, coverage disputes, or subrogation.
Portfolio-level data for cedents
Aggregate audit signals across the AI portfolio you cede. Munich Re, Swiss Re, and Lloyd's syndicate cedents need quantitative risk data to take on AI exposure — your audit record set gives them what no other source can today.
Evidence-based risk assessment by design
State insurance commissioners and EIOPA require evidence-based risk assessment for emerging risk categories. Audit records are designed to fit those documentation requirements — reduce friction in new-product filings.
Built for regulated finance
Documentation that fits your existing compliance framework
Audit outputs are designed to feed Solvency II classification, IDD demands-and-needs, AI Act high-risk filings, and DORA third-party documentation. No bolt-on; integrates.
How we engage
Three ways to start
30-minute walkthrough
A briefing for your underwriting + product team. How the audit works, what comes back, how it plugs into your existing workflow. No commitment.
Run it on your book
Audit 5-25 models from your existing AI insurance pipeline. We provide the methodology + audit results; you keep all customer data and underwriting decisions.
Integrated underwriting
Audit infrastructure embedded in your underwriting workflow. Per-audit fee or annual programme licence depending on volume. Available via fronting arrangement until our own MGA authorisation is in place.
Who you're working with
Sweden-based, EU/EEA-aligned, regulator-friendly
The model.insure programme is operated by Droidtech 42 AI Labs AB (Sweden, org.nr 559534-0745). All commercial conversations and KYC review today are with Droidtech AB. An EU/EEA insurance distribution authorisation is being pursued; intermediate carrier partnerships are available via fronting arrangements with licensed EU/EEA carriers.
Talk to us before your next AI policy goes out.
30-minute briefing with the team. No deck-trade, no sales script — engineers + underwriter conversation.