Open-source policy diffusion observatory

Propagate

How tech policy spreads, changes, and gets captured.

Technology policy in the U.S. is made state by state, but it doesn't originate state by state. Bills travel โ€” borrowing text, weakening provisions, absorbing industry amendments, and arriving in the tenth state looking very different from the first. Propagate makes that process visible. It treats policy diffusion as an analytical object, not a background condition.

Policy Families

7 domains tracked for MVP
Algorithmic Discrimination / ADMTLive
8 bills ยท 5 states
Bills introduced per year
20222023202420252026
Consumer AI Transparency / Chatbot DisclosurePlanned
Bills requiring disclosure when consumers interact with AI systems, including chatbot labeling and synthetic media provenance.
Data Center Siting & EnergyPlanned
State-level bills regulating the siting, energy consumption, and water use of large-scale data center facilities.
Autonomous VehiclesPlanned
State frameworks governing the testing, deployment, and liability of autonomous and semi-autonomous vehicles.
Social Media & MinorsPlanned
Age verification, parental consent, and design-safety bills targeting social media platforms and minors.
Frontier / Foundation Model RegulationPlanned
Bills imposing safety evaluations, liability, or registration requirements on large-scale AI foundation models.
AI-Specific Data PrivacyPlanned
Privacy frameworks specifically addressing AI training data, biometric data, and inferential data uses.

About

Propagate is an open-source research tool for tracking how technology policy spreads across U.S. states. It maps bill genealogies, measures diffusion speed, records organized opposition, and makes the underlying data available for download. Primary audiences are policy researchers, investigative journalists, legislative staff, and public interest advocates.

Data coverage notice: Propagate is in early development. The ADMT / Algorithmic Discrimination family is seeded with verified data (8 bills, 5 states). All other policy families are planned for future contributions. Lobbying and advocacy data is incomplete by design and permanently under construction. All data points link to primary sources. Read the methodology.