Americans for Responsible Innovation Is an AI Policy Lobby

Americans for Responsible Innovation (ARI) is a Washington, D.C., group that lobbies on how the U.S. government should regulate artificial intelligence (AI). It holds a 501(c)(4) tax status, which lets it lobby and does not require it to disclose donors.

What Americans for Responsible Innovation Is and Who Pays for It

ARI is a Washington, D.C., nonprofit that lobbies Congress and federal agencies on AI rules. Its 2024 fiscal year brought in $5,244,469 in total revenue, according to Internal Revenue Service (IRS) filing data published by ProPublica. The group names its donors on its own About page, even though its 501(c)(4) status doesn't require that.

ARI says it works to help U.S. policymakers build "a thoughtful governance framework for rapidly advancing technologies." It says it was founded to be nonpartisan and independent of the technology industry. Its About page lists ten core principles, including regulating according to risk and avoiding regulatory capture.

What the Filings Show

The filings reviewed here differ sharply. In 2023, ARI reported $501,673 in revenue and $23,064 in expenses. In 2024, expenses rose to $2,592,128, and total assets reached $3,514,963.

Revenue exceeded expenses by $2,652,341 in 2024. ProPublica notes that its summary data may be incomplete or out of date, so every figure here carries its filing year.

Who Funds ARI

The About page names three institutional supporters: Inclusive Abundance Action, Omidyar Network and Coefficient Giving. Coefficient Giving is the renamed Open Philanthropy. The grant amount couldn't be confirmed from the grant page, so none appears here.

The named individual donors are Brad and Julie Carson, Eric Gastfriend and Caroline Mehl, and Steve Newman. Two of those individual donor entries belong to the group's own leaders. Brad Carson is its president, and Eric Gastfriend is its executive director.

ARI says it doesn't take money from for-profit corporations or non-U.S. entities, and that its donors don't control its operations. Founders Pledge recommends ARI for giving, according to the About page.

How Much ARI Spends on Lobbying

Legis1 reports that ARI has spent $2.81 million on federal lobbying since April 2024. That report is dated 14 March 2026. Legis1 links to its own platform pages rather than to the underlying federal lobbying filings, and those filings couldn't be read for this research.

Legis1 names three outside firms. Kountoupes Denham Carr & Reid LLC was listed at $600,000, Mindset Advocacy LLC at $240,000, and Navigators Global LLC at $140,000.

Legis1 also reports that ARI retained Neil Hill Global Strategies LLC on 9 January 2026, with no amount given. ARI's federal lobbying registration lists AI regulation as an issue area, covering science and technology, the computer industry and consumer protection. Some annual lobbying totals circulating on a third-party wiki weren't confirmed, so none appear here.

Its Political Action Committee

ARI also has a separate political action committee (PAC), Americans For Responsible Innovation Ltd. Political Action Committee, with Federal Election Commission (FEC) committee ID C00893883. Its campaign-finance activity is distinct from the group's lobbying. Its receipts and spending aren't published here, so no PAC dollar figure appears.

What ARI Works On: Bills and Agency Processes

ARI's central fight is against federal preemption of state AI laws. Brad Carson, ARI's president, has called preemption a "new Section 230" in a Senate letter ARI led.

The group's stance sits inside the wider debate over AI rules. For that background, see the AI regulation explainer. ARI also says it addresses current AI harms, national security concerns and emerging risks.

In 2025, ARI helped defeat proposals to block state AI laws. Wikipedia's account lists its tactics: a website that collected anti-preemption statements, a press conference with state lawmakers, and a grassroots petition against a moratorium on state AI laws. For the state side of that split, see the state vs federal AI regulation explainer.

On 3 December 2025, House Majority Leader Steve Scalise said Congress would not include state AI-law preemption in the fiscal 2026 National Defense Authorization Act (NDAA). ARI's own post, which is a self-report, credits a bipartisan coalition of state lawmakers, parents, faith leaders, unions and whistleblowers. Carson said Congress has now "gotten the message."

An earlier attempt to attach AI-law preemption to H.R. 1, which the sources call the "Big Beautiful Bill," failed. ARI also announced an ad campaign in Massachusetts urging Rep. Lori Trahan (MA-03) to oppose a ban on state AI legislation.

The Senate letter ARI led with the Institute for Family Studies was signed by 28 organizations. Signers include Common Sense Media, Issue One, Young People's Alliance, Encode AI and Fairplay. The letter was framed around the anniversary of Section 230. For how AI bills move through Congress, see the AI legislation explainer.

The table lists the bills and proposals in ARI's record, with the outcome where the sources give one.

The TAKE IT DOWN Act requires platforms to remove non-consensual deepfakes and intimate images after user requests, and ARI supported it. The act passed in April 2025, according to Wikipedia's citations.

ARI supports export controls on advanced AI chips. In September 2025, an ARI spokesperson said the GAIN AI Act would "take back some control over what technology we're exporting to China." The sources here don't establish that bill's outcome.

ARI also organized a letter signed by 40 economists in September 2025. The letter asked the federal government to collect more data on AI's workforce impact.

On federal agencies, ARI advocates modernizing and funding agency AI capacity rather than only restricting industry. Wikipedia dates its Commerce Department advocacy to 2024 but cites a 12 May 2025 report, so the dates on that page don't line up.

ARI also takes aim at industry lobbying. An ARI official, quoted by Legis1, called Andreessen Horowitz "an absolutely powerful lobbying force, possibly the most powerful single organization I've seen in recent years." The sources don't name that official. ARI opposes the firm's efforts to weaken state AI laws.

For the executive branch side of AI policy, see the Trump AI executive orders explainer. For other groups in the fight, see who is fighting AI.

Who Runs ARI

Brad Carson is ARI's president. The About page and Wikipedia both describe him as a former U.S. congressman and Department of Defense official. Eric Gastfriend is its executive director, and both sources describe him as a technology entrepreneur.

Wikipedia says Carson and Gastfriend launched ARI in March 2024, citing reporting dated 6 March 2024. ARI's own About page says it was founded in late 2023 and doesn't give an exact date. The safest reading is that ARI was set up in late 2023 and publicly launched in early 2024.

Douglass Calidas is senior vice president of government affairs. On the 2024 filing data, he was the highest-compensated officer, at $438,631. The 2023 data lists Gastfriend at $0.

ARI's About page lists principles including navigating uncertainty, maintaining America's lead and adopting AI in government responsibly. A job posting from the organization describes its approach as bipartisan, building coalitions across the political spectrum.

No other officers or board members are named in the filings reviewed.

Why Its Legal Form Matters

ARI is a 501(c)(4) nonprofit, and that status permits it to lobby. The same status doesn't require it to disclose donors. The donor list therefore depends on the group's own choice to publish it.

Its exact legal name on IRS filings is Americans For Responsible Innovation Ltd. Its employer identification number (EIN) is 93-3248564, and it is filed in the District of Columbia. ProPublica's Nonprofit Explorer lists its subsection as 501(c)(4).

ProPublica also notes that ARI doesn't appear in the IRS's most recent list of tax-exempt organizations. That is why every figure above carries its filing year.

Whether ARI runs a separate 501(c)(3) research arm isn't confirmed in the filings. To check the filings yourself, start with the ProPublica record for EIN 93-3248564.

The Criticism of ARI and What Is Unconfirmed

No critic of ARI is on the record by name in the sources below. The critiques below come from the structure of its funding, its own filing numbers and reporting by Legis1. None is a quote from a named critic.

The strongest criticism the sources support is structural. A 501(c)(4) doesn't have to disclose its donors, so ARI's donor list is voluntary and may be incomplete. Two of the named individual donors are its own leaders, so the people who run the group also help fund it.

The second point is the gap between money raised and money spent. In 2024, ARI reported $5,244,469 in revenue and $2,592,128 in expenses, so expenses were about 49 percent of revenue. The same filing shows $438,631 paid to its top government-affairs officer. A reader can judge that ratio for themselves.

Legis1 raises a third point. It reports that Jonathan Hale and Narda Jones, the lobbyists on the Neil Hill engagement, are making their federal lobbying debuts with no prior congressional experience. The sources reviewed don't include a response from ARI on that point.

The fourth point concerns independence. ARI says it is nonpartisan and independent of the technology industry. Its institutional supporters include Omidyar Network and Coefficient Giving, which are both technology-fortune philanthropies. Readers can set the claim beside the funder list.

Industry groups argue that federal preemption prevents a 50-state patchwork of AI rules. ARI disputes that position, according to Legis1.

A third-party wiki describes ARI as backed by donors aligned with effective altruism. That wiki is compiled rather than official, and its verification notes for related figures were marked unverified. None of that framing appears here as fact.

What the Sources Don't Settle

• The founding month is contested. ARI's About page says late 2023, while Wikipedia says March 2024.
• The date and address of ARI's federal lobbying registration aren't confirmed, because the federal lobbying filings couldn't be read for this research.
• The amount of the Coefficient Giving grant isn't confirmed, because the grant page couldn't be read. No figure is published here.
• Outcomes aren't established for the Future of AI Innovation Act, the CREATE AI Act, the GAIN AI Act, the AI Workforce PREPARE Act and the Economy of the Future Commission Act.
• Wikipedia describes a letter from 20 national security experts urging limits on Nvidia H20 sales to China, and a proposal that AI labs pass safety review for government contracts. The page doesn't clearly tie either to ARI, so neither is attributed to ARI here.
• No 2025 or 2026 financial filing is available on ProPublica, so 2024 is the latest revenue figure.
• The PAC's receipts and spending weren't read, so no PAC figures appear here.

To follow the preemption fight further, read the state vs federal AI regulation explainer next.

Sources

• ARI, About page
• ProPublica Nonprofit Explorer, Americans For Responsible Innovation Ltd
• Wikipedia, Americans for Responsible Innovation
• ARI, AI law preemption dropped from the NDAA
• ARI, Senate letter on the Section 230 anniversary
• ARI, Massachusetts ad campaign
• Legis1, AI regulation lobbying expansion (14 March 2026)
• ProPublica Itemizer, ARI PAC statement of organization
• Open Philanthropy (now Coefficient Giving), grant page for ARI
• The Impact Job, ARI job posting
• LongTermWiki, Americans for Responsible Innovation (third-party, unverified)