Policy Entrepreneurship with AI
Instructor information
Dr. José Marichal (he/him/his)
Professor of Political Science
Contact: marichal@callutheran.edu
Course Description: The AI Policy Entrepreneur
The AI Policy Entrepreneur is emerging at the intersection of technology, advocacy, and governance. They work in think tanks, advocacy organizations, lobbying firms, and increasingly within government itself. Unlike the APO focused on electoral campaigns, the AI Policy Entrepreneur is focused on policy change — getting legislation passed, regulations shifted, public programs funded, or government priorities reordered.
This course prepares you for this role. We explore how AI doesn't just speed up old processes, but fundamentally changes what is possible at each stage of the policy cycle, from agenda setting to termination.
Course Objectives
By the end of this course, students will be able to:
- Deploy AI for Issue Monitoring: Systematically track legislative and regulatory activity across multiple jurisdictions.
- Draft Policy with AI Assistance: Use LLMs to generate legislative language, regulatory comments, and policy briefs.
- Simulate Policy Impact: Model the effects of policy decisions on different stakeholders.
- Navigate Bureaucracy: Use AI to understand agency procedures and identify key decision points.
- Evaluate Ethics: Critically assess the democratic implications of automated advocacy and influence.
Course Assignments
Total Points: 100
- Issue Monitoring Dashboard (20 Points): Build an automated system to track a specific policy issue across 50 states and federal agencies (using tools like automated keyword scanning or API integration).
- Legislative Drafting Portfolio (25 Points): Retrieve a model policy and use AI to draft three variants: a tax credit, a grant program, and a regulatory mandate. Includes an analysis of the "Goldilocks zone" for coalition support.
- Regulatory Comment Bot (25 Points): Design a system (conceptually or functionally) to generate unique, substantive comments for a mock rulemaking process. Includes an ethical reflection on "astroturfing" vs. legitimate advocacy.
- Final Policy Simulation (30 Points): A semester-long project where you act as a Policy Entrepreneur for a cause, moving it through all six stages using AI tools.
Schedule
Unit 1: The New Class of Advocate
Introduction to the AI Policy Entrepreneur and the tools of the trade.
How the old guard of lobbying is being challenged by data-driven scale. The shift from cocktail parties to server farms.
Social listening AI (Brandwatch), Legislative AI (Plural), and predictive analytics. Setting up your environments.
Unit 2: Stage 1 — Agenda Setting
Making problems visible.
Using AI to scan legislative activity and regulatory filings. Flagging "weak signals" before they become mainstream news.
Generating localized narratives for different audiences (rural vs. urban). Using AI sentiment analysis to test which frames resonate.
Processing open government data to build interactive visualizations that localize national issues and drive agenda setting. Case Study: Save Our Bridges, an interactive map using Federal Highway Administration data to help citizens and businesses seamlessly locate structurally deficient and fracture-critical bridges in their zip codes.
Unit 3: Stage 2 — Policy Formulation
Designing solutions that survive.
Ingesting policies from other jurisdictions to see what passed and what failed. "This worked in Colorado, but failed in Texas."
Generating legislative variants (tax credit vs. mandate) in hours. Using AI to red-team your own legislation for loopholes.
Unit 4: Stage 3 — Decision-Making
Building coalitions and winning votes.
Building deep profiles of decision-makers: voting history, donors, and committee leverage. Identifying who is persuadable.
Generating unique pitch materials for every member of a committee. The Ag Chair gets the rural angle; the Tech Caucus gets the innovation story.
Preparing for the gauntlet. Using AI to simulate hostile questioning from opposition legislators based on their past behavior.
Unit 5: Stage 4 — Implementation
Ensuring follow-through (where most policies die).
Generating thousands of substantive comments during rulemaking. The ethics of synthetic public participation.
Using AI to help entities understand new rules (reducing resistance). Monitoring agencies for enforcement gaps.
Unit 6: Stage 5 — Evaluation
Measuring success and learning.
Collecting administrative data, FOIA responses, and budget documents to build datasets on policy outcomes.
Using synthetic controls and difference-in-differences to estimate effects. Producing automated reports for stakeholders.
Unit 7: Stage 6 — The Long Game
Maintenance, Expansion, or Termination.
Monitoring for repeal efforts and budget cuts. Keeping the coalition alive with automated updates.
Identifying new jurisdictions for successful policies. Or, using AI to build the case for killing a failed policy.
"We're not just making policy faster. We're changing who gets to make it." The accountability gap and the future of governance.