Americans for Responsible Innovation

Americans for Responsible Innovation (ARI) is a Washington-based bipartisan nonprofit launched in March 2024 to advocate for federal AI governance that protects the public without sacrificing American competitiveness. Operating as a 501(c)(4) independent of for-profit corporations and non-US entities, ARI combines Washington policy expertise, coalition organizing, and direct voter mobilization — assembling advocacy coalitions around specific legislation, gathering grassroots petitions, and routing survivor testimony into the congressional record — on a range of AI consumer-safety, national-security, and emerging-risk issues.

Founding and structure

ARI was co-founded by Brad Carson — former Oklahoma Congressman, Under Secretary of the Army, and Acting Under Secretary of Defense — and Eric Gastfriend, former CEO of DynamiCare Health, a digital health startup that received awards from the NIH and FDA. The two described the organization's purpose as filling a policy-advocacy gap: in 2024, major AI companies were heavily engaged in Washington, while independent civil-society capacity for AI governance advocacy was thin. ARI's declared independence from industry funding — no money from for-profit corporations or non-US entities — and its bipartisan positioning (Carson a Democrat, the founding board including Republicans and libertarians, the board of advisors including former Republican Congressman Tim Ryan) were designed to give its advocacy credibility across party lines.

The organization maintains a political action committee, ARI PAC, for electoral-adjacent activities. Its documented philanthropic funders include Omidyar Network, Inclusive Abundance Action, and Coefficient Giving. The board of advisors includes Scott Aaronson (UT Austin), Erik Brynjolfsson (Stanford Digital Economy Lab), Gillian Hadfield (Johns Hopkins, former OpenAI policy advisor), Stuart Russell (UC Berkeley), and Samuel Hammond (Foundation for American Innovation). ARI's ten stated governing principles include navigating regulatory uncertainty, building durable regulatory capacity, avoiding regulatory capture, and moving quickly — an explicit rejection of the argument that AI governance should wait for technical consensus to form.

TAKE IT DOWN Act

ARI's most prominent early legislative win was the TAKE IT DOWN Act, signed by President Trump in May 2025 — the first major federal law governing generative AI harms. The law criminalises the publication of non-consensual sexually exploitative images, including AI-generated deepfakes, and requires online platforms to implement notice-and-takedown processes for such content. ARI described it as the most significant federal AI safety law enacted in the United States.

ARI's campaign assembled a coalition of 30+ children's safety and tech-policy organisations, organised legal scholars to provide constitutional analysis through a memo and press conference, gathered thousands of voter signatures supporting the legislation, and coordinated voter calls to legislators' offices. The campaign provided a national platform for Megan Garcia — the mother of Sewell Setzer, a 14-year-old who died by suicide after an extended relationship with an AI companion chatbot — whose testimony at congressional hearings became the evidentiary anchor for the bill's passage. The combination of institutional coalition-building and survivor-led testimony demonstrated ARI's operating method: DC policy machinery and first-person human cost in a single campaign.

Anti-preemption campaign

In 2025, ARI led the effort to defeat a proposed 10-year federal moratorium on state and local AI regulations buried in a budget reconciliation bill — a provision backed by major technology companies that would have nullified the significant body of state-level AI consumer-protection law enacted between 2023 and 2025.

ARI, alongside coalition partners Accountable Tech and Public Citizen, launched DefendAISafeguards.org on June 4, 2025, which gathered over 25,000 grassroots petition signatures in its first two weeks. The organisation simultaneously coordinated a letter signed by 260 bipartisan state lawmakers opposing the moratorium and published original research documenting the consequences of preempting state AI law. The Senate voted 99-1 to strike the moratorium provision from the reconciliation bill. ARI characterised the outcome as proof that bipartisan consensus on AI consumer protection is achievable when public pressure and coalition coordination are sustained.