$ status --board  →  recruiting founding members  ·  get involved
// policy · civic ai accountability standard

We build the rules that should exist for AI.

And we take them directly to the people who write the laws.

/*tl;dr*CAAS is a public-domain review process for AI systems, modeled on environmental impact review.*We also wrote a model law, the Civic AI Accountability Act of 2026.*The full packet has been delivered to four congressional offices.*/
// 01 the problem

AI is being deployed without oversight.

When a company wants to build a new factory, it has to complete an environmental impact review. When a technology company deploys a new AI system that affects millions of people, there is no equivalent process. No review. No public input. No assessment of who might be harmed.

Neurodivergent people, elderly users, and disabled individuals are often the most affected and the least consulted.

// 02 our solution

The Civic AI Accountability Standard.

/*tl;dr*A structured way to evaluate AI systems before they launch and hold companies accountable afterward. Two instruments, one diagnostic, all public domain.*/
PUBLIC DOMAIN FRAMEWORK

CAAS adapts the logic of environmental impact review to AI deployment. It gives legislators, regulators, and advocacy organizations a structured way to evaluate AI systems before they launch, and to hold companies accountable after the fact.

ACIR
Algorithmic Civic Impact Review

A pre-deployment checklist evaluating systems across seven areas, including effects on vulnerable populations.

CTIR
Civic Technology Integrity Review

A post-deployment tool for congressional staff, regulators, and attorneys to evaluate whether a system does what was promised.

8Q
Social capital diagnostic

Eight questions measuring whether a platform builds community trust or erodes it.

license: public domain · anyone can use, adapt, or cite it without restrictionRead the full framework at CAASNow.org
// 03 model law

The Civic AI Accountability Act of 2026.

We did not stop at a framework. We wrote a model law any state or federal legislator can introduce. It would require civic impact review before deploying large-scale AI systems, mandate assessment of effects on vulnerable populations, and create enforcement mechanisms for post-deployment audits.

license: free for any legislative office to use or adapt
// 04 who we have contacted

This work is not sitting on a shelf.

The complete CAAS research packet has been formally delivered to four congressional offices.

Senator Tammy BaldwinD-WI · delivered
Senator Bernie SandersI-VT · delivered
Senator Elizabeth WarrenD-MA · delivered
Representative Alexandria Ocasio-CortezD-NY-14 · delivered

Our founder has an existing relationship with Senator Baldwin's office, which has previously acted on his behalf in a congressional inquiry with the Social Security Administration.

// 05 consumers

Fighting back on behalf of consumers.

When companies harm neurodivergent consumers, we take direct action: formal complaints, demand letters, and detailed research documenting unfair billing practices, cancellation barriers, misleading marketing, data privacy violations, and inadequate diagnostic services. Each case serves the individual affected and contributes to the public record.

Questions about our policy work and media inquiries are welcome via the Contact page.