Risk AssessmentHigh riskComparison recommended

AI Playbook for Environmental Liability Underwriting

A pollution liability underwriter has received a submission from a dry cleaner chain with 22 locations. Three locations have known PCE (perchloroethylene) contamination from prior operators. The applicant is requesting site-specific coverage for the 3 contaminated sites and blanket coverage for the other 19.

When to use this playbook

  • Use this playbook when the decision looks like the situation above: A pollution liability underwriter has received a submission from a dry cleaner chain with 22 locations.
  • It is a fit when you have source files in hand and need a structured, reviewable analysis — not a generic chat answer about "Environmental Liability Underwriting".
  • Do not use it as a substitute for licensed, legal, clinical, or authorized official judgment in the domain.

What you'll need

  • Application with site descriptions and environmental history
  • Phase II ESA reports for the 3 contaminated sites
  • Regulatory correspondence for the contaminated sites (remediation status)
  • Dry cleaner industry pollution loss data
  • Prior carrier loss run (if available)

Attachments: Multiple attachments (Spreadsheets, Documents)

The Prompt

You are a pollution liability underwriter evaluating a dry cleaner chain with legacy contamination. I am attaching:

Work only from the attached source files. If a conclusion is not supported, say so.

Produce:
1. Assess the remediation status and liability exposure for each of the 3 contaminated sites: how much cleanup cost is remaining, and what is the third-party liability exposure?
2. Evaluate whether the known contamination is insurable under a site pollution liability policy or whether it requires a remediation cost cap structure.
3. Assess the 19 non-contaminated sites: what is the frequency and severity exposure for a dry cleaner blanket pollution program?
4. Calculate the total premium indication for the structure requested: site-specific for 3 contaminated + blanket for 19 clean sites.
5. Tell me the underwriting decision: write as submitted, restructure the coverage (separate policies, different limits/retentions), or decline the contaminated sites.

Call out where independent models are likely to disagree, and list follow-up documents a reviewer should request.

What to expect

  • Contaminated site liability exposure assessment
  • Insurability analysis for known contamination
  • Non-contaminated site frequency/severity analysis
  • Total premium indication with structure
  • Write/restructure/decline recommendation with rationale

Review before you act

  • Validate this output against source files before relying on it: Assess the remediation status and liability exposure for each of the 3 contaminated sites: how much cleanup cost is remaining, and what is the third-party liability exposure?.
  • Validate this output against source files before relying on it: Evaluate whether the known contamination is insurable under a site pollution liability policy or whether it requires a remediation cost cap structure.
  • Validate this output against source files before relying on it: Assess the 19 non-contaminated sites: what is the frequency and severity exposure for a dry cleaner blanket pollution program?.
  • Validate this output against source files before relying on it: Calculate the total premium indication for the structure requested: site-specific for 3 contaminated + blanket for 19 clean sites.
  • Confirm every cited figure, date, counterparty, or requirement against the attached originals — models compress and can drop a qualifier.
  • Treat disagreement between models as a review item, especially on classification, materiality, and recommended next action.
  • Do not authorize an operational, clinical, legal, credit, or enforcement action solely because the models agree.

Why compare models on this

For Environmental Liability Underwriting, running the same attachments across independent models is useful because the hard part is classification and completeness, not fluency. The workflow is already designed to surface contaminated site liability exposure assessment; insurability analysis for known contamination; non-contaminated site frequency/severity analysis; total premium indication with structure. Those are comparison artifacts — they only exist if more than one model runs. Models split on tail scenarios, aggregation, and whether a hazard is excluded. Divergence is a referral to a specialist underwriter, not a silent average of three prices.

Insurance UnderwritingSpecialty LiabilityRisk AssessmentHighMultiple attachments

See governed multi-model AI on your own prompt

Compare GPT-5, Claude, and Gemini side by side, with human review and a decision record built in.