AI Product Liability Exposure Assessment Playbook
A specialty lines underwriter has received an application from a $120M revenue medical device company seeking $10M/$20M product liability limits. The company's flagship device has been in market for 3 years with no recalls but has 12 pending product liability claims, 4 of which involve alleged serious injury.
When to use this playbook
- Use this playbook when the decision looks like the situation above: A specialty lines underwriter has received an application from a $120M revenue medical device company seeking $10M/$20M product liability limits.
- It is a fit when you have source files in hand and need a structured, reviewable analysis — not a generic chat answer about "Product Liability Exposure Assessment".
- Do not use it as a substitute for licensed, legal, clinical, or authorized official judgment in the domain.
What you'll need
- Product liability application and loss run (3 years + pending claims list)
- Device description, FDA clearance documentation, and market volume
- 4 serious injury claim summaries (redacted)
- Industry product liability benchmarks for Class II medical devices
- FDA MAUDE database search results for the device
Attachments: Spreadsheets (Spreadsheets)
The Prompt
You are a product liability underwriter evaluating a medical device company with adverse claims trends. I am attaching: Work only from the attached source files. If a conclusion is not supported, say so. Produce: 1. Assess the claims development pattern: are the 12 pending claims a leading indicator of a mass tort, or are they consistent with normal product liability frequency for a device of this type? 2. Analyze the 4 serious injury claims: what is the injury mechanism, is there a common failure mode, and does the MAUDE data show similar reported injuries? 3. Calculate the loss development exposure: what is the estimated ultimate claims cost if the pending claims develop adversely? 4. Identify whether the claimed limits ($10M/$20M) are adequate for the exposure and what the appropriate retention and layer structure should be. 5. Tell me the underwriting decision and the terms, conditions, and exclusions that adequately protect the carrier. Call out where independent models are likely to disagree, and list follow-up documents a reviewer should request.
What to expect
- Claims development pattern vs. mass tort indicators
- Serious injury mechanism and common failure mode analysis
- Ultimate claims cost projection
- Limit and retention adequacy assessment
- Underwriting decision with terms, conditions, and exclusions
Review before you act
- Validate this output against source files before relying on it: Assess the claims development pattern: are the 12 pending claims a leading indicator of a mass tort, or are they consistent with normal product liability frequency for a device of this type?.
- Validate this output against source files before relying on it: Analyze the 4 serious injury claims: what is the injury mechanism, is there a common failure mode, and does the MAUDE data show similar reported injuries?.
- Validate this output against source files before relying on it: Calculate the loss development exposure: what is the estimated ultimate claims cost if the pending claims develop adversely?.
- Validate this output against source files before relying on it: Identify whether the claimed limits ($10M/$20M) are adequate for the exposure and what the appropriate retention and layer structure should be.
- 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 Product Liability Exposure Assessment, 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 claims development pattern vs. mass tort indicators; serious injury mechanism and common failure mode analysis; ultimate claims cost projection; limit and retention adequacy assessment. 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.
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.

