Sen. Mark Kelly · 2025

Sen. Mark Kelly — "AI for America" Roadmap

Kelly

A September 2025 roadmap organized around a single funding question most AI roadmaps skip: who pays? Kelly's answer is the AI Horizon Fund, a federal trust fund financed by revenue from the largest frontier AI companies and spent through mandatory spending outside the annual appropriations process. Candidate bases are floated rather than chosen — large-scale use of public resources such as power, water, and land, profits from AI-powered digital ad tools, or AI revenue windfalls. The Fund pays for worker upskilling through union apprenticeships and community colleges, an AI economic adjustment program with more generous unemployment insurance, and energy and water infrastructure. A full chapter on data centers holds that they must pay for the grid assets they require rather than leaving stranded costs to ratepayers, and calls for disclosure standards on energy, water, and emissions. A shorter trust chapter says models must be rigorously tested before release, including red-teaming by trusted entities such as government agencies and national labs supported by the Fund.

Key Provisions

Regulatory Philosophy

Government-facilitated redistribution and proactive workforce adjustment. Kelly's approach treats AI primarily as an economic disruption problem requiring institutional investment, not just a safety or innovation problem. His emphasis on industry-funded transition mechanisms rather than taxpayer-funded programs distinguishes him from both the White House and progressive approaches.

Where the burden falls

FrontierUnspecifiedAppsUnspecifiedHyperscalersUnspecifiedChipsUnspecified
Base assessed
Unchosen — the Horizon Fund names a payer and three candidate bases, but picks none
Why it lands there
The Horizon Fund is aimed at 'the biggest frontier companies,' but the roadmap offers three candidate bases and picks none, and each lands on a different layer. A levy on large-scale use of power, water, and land assesses whoever owns the data centers, which is the hyperscalers. A levy on profits from AI-powered ad tools reaches the consumer platforms — the app layer's oldest business — and passes over the labs. A windfall levy on AI revenue is the only one that lands squarely on the frontier. The choice of base is the whole policy, and it is still open — which makes this the proposal most worth pressing a sponsor on.
Compare all proposals by layer →

Strengths

Derived from the proposal’s own policy documents

  • +The AI Horizon Fund creates a concrete, industry-funded mechanism for worker transition rather than leaving displaced workers to fend for themselves
  • +Advance notice requirements for AI-driven workforce changes give workers and communities time to prepare, modeled on proven WARN Act principles
  • +BLS data collection expansion would finally give policymakers real numbers on AI-driven job creation, displacement, and retraining outcomes
  • +Bipartisan security provisions with Sen. Young demonstrate the proposal can attract cross-aisle support
  • +Industry-funded rather than taxpayer-funded transition mechanisms place costs on the companies profiting from disruption

Weaknesses

From the perspective of political opposition

  • −The Horizon Fund's financing is a list of possibilities rather than a mechanism — use of power, water, and land, AI-powered ad profits, or AI windfalls are offered as 'a starting point for conversation', so the roadmap commits to a payer without committing to a tax
  • −Safety gets two pages: pre-release testing and red-teaming are endorsed without saying who requires them, what happens if a model fails, or which models are covered, and catastrophic risk is not addressed — child safety, election misuse, and liability are posed as open questions rather than answered
  • −No enforcement mechanism for workforce change notifications — companies can simply ignore advance notice requirements the way they routinely ignore existing labor protections
  • −Frames AI disruption as a manageable adjustment problem when the scale of displacement could be civilizational, offering band-aids for bullet wounds
  • −The bipartisan framing is a way to avoid taking hard positions on preemption, liability, or any issue where industry might push back

On the Podcasts

Verbatim from podcast transcripts; each links to the moment it was said

“I wrote a paper called AI for America, and what I'd like to do is form this thing called the AI Horizon Fund, where these corporations that are building this artificial intelligence, these systems are contributing money into this fund. There's ways we can do this. I've talked to the companies about it. Some of them are supportive because they understand some of the challenges that we're going to face as a nation.”
Mark Kelly, U.S. Senator (D-AZ), AI for America authorThe Nicole Sandler Show · Sen. Mark Kelly Changed My Mind, Something I Didn't Expect · Jun 8, 2026
“Senator Mark Kelly has been interesting… he has taken a little more of a middle ground approach of sort of embracing the technology while making sure it can still empower Americans.”
Ashley Gold, Semafor senior technology correspondentThe Odd Years · Both parties' voters are freaked out by AI. So what is Congress going to do about it? · Jun 16, 2026
“What I'm torn about with this is every other government program that's supposed to be helping people doesn't. Look at the waste and the fraud and the abuse and the red tape and the nonsense… I think there's probably a better way of doing this than having some kind of a fund.”
Mike Broomhead, KTAR News Phoenix radio hostThe Mike Broomhead Show · Hour 4: Golden Ticket · Sep 23, 2026

Position on Analytical Frameworks

Enforcement Mechanism vs. Regulatory Scope

Prevention vs. Liability & Regulatory Authority

Innovation Priority vs. Worker Protection

Pre-deployment Obligations vs. Federal Preemption

← Back to all proposals