AI Regulatory Impact Analysis Playbook
A federal agency is developing a proposed rule requiring new safety standards for industrial facilities. The rule affects approximately 4,200 facilities. The agency's economists estimate compliance costs of $2.4B over 10 years. OMB review is required under Executive Order 12866.
When to use this playbook
- Use this playbook when the decision looks like the situation above: A federal agency is developing a proposed rule requiring new safety standards for industrial facilities.
- It is a fit when you have source files in hand and need a structured, reviewable analysis — not a generic chat answer about "Regulatory Impact Analysis".
- Do not use it as a substitute for licensed, legal, clinical, or authorized official judgment in the domain.
What you'll need
- Proposed rule text and agency technical analysis
- Cost estimate methodology and facility survey data
- Injury and fatality data for the affected industry (past 10 years)
- EPA and OSHA value of statistical life (VSL) guidance
- OMB Circular A-4 requirements
Attachments: Documents (Documents)
The Prompt
You are a regulatory economist developing the Regulatory Impact Analysis for a federal safety rule affecting 4,200 facilities. I am attaching: Work only from the attached source files. If a conclusion is not supported, say so. Produce: 1. Structure the benefit-cost analysis: total compliance costs ($2.4B) vs. quantified benefits (lives saved x VSL, injuries prevented). 2. Assess the uncertainty in both estimates and develop the sensitivity analysis OMB will require. 3. Identify the distributional effects: which facility types and regions bear the highest compliance costs? 4. Assess the alternatives analysis: has the agency considered less costly alternatives achieving similar safety outcomes? 5. Tell me whether the RIA will survive OIRA review and what the most likely challenge points are. Call out where independent models are likely to disagree, and list follow-up documents a reviewer should request.
What to expect
- Benefit-cost analysis with quantified benefits
- Sensitivity analysis framework
- Distributional effects analysis
- Alternatives analysis assessment
- OIRA survivability prediction and challenge points
Review before you act
- Validate this output against source files before relying on it: Structure the benefit-cost analysis: total compliance costs ($2.4B) vs. quantified benefits (lives saved x VSL, injuries prevented).
- Validate this output against source files before relying on it: Assess the uncertainty in both estimates and develop the sensitivity analysis OMB will require.
- Validate this output against source files before relying on it: Identify the distributional effects: which facility types and regions bear the highest compliance costs?.
- Validate this output against source files before relying on it: Assess the alternatives analysis: has the agency considered less costly alternatives achieving similar safety outcomes?.
- 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 Regulatory Impact Analysis, 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 benefit-cost analysis with quantified benefits; sensitivity analysis framework; distributional effects analysis; alternatives analysis assessment. Those are comparison artifacts — they only exist if more than one model runs. Models disagree on exemption applicability, IG finding risk, and how aggressive a budget narrative can be. Those disagreements belong with counsel and the authorizing official.
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.

