How much have AI super PACs spent on the 2026 midterms?

AI-aligned super PAC spending in the 2026 US midterms has now topped $185 million according to Yahoo News, Gadget Review, and Guardian reporting from June 22-23, 2026. The OpenAI-backed super PAC has reportedly spent ~$23.5 million; Anthropic-backed groups ~$16.6 million. The remainder comes from broader tech-aligned and AI-industry-aligned groups operating in the same lane. Roughly half of the AI-PAC spending is concentrated on a single New York City congressional primary race involving state assemblyman Alex Bores, who sponsored New York's RAISE Act AI safety legislation.

What is the RAISE Act and why are AI companies fighting over it?

The RAISE Act is New York state legislation co-sponsored by Assemblyman Alex Bores that would impose AI safety obligations on frontier model developers — including third-party safety assessments, incident disclosure, and liability provisions for catastrophic harms. Anthropic-aligned groups have broadly supported RAISE Act-style legislation as aligned with Anthropic's responsible-scaling-policy stance. OpenAI-backed groups argue that state-by-state AI rules would create 'a chaotic patchwork of state rules that would crush innovation' — preferring federal pre-emption. The NYC primary race involving Bores is the immediate political battleground.

Why are OpenAI and Anthropic spending against each other?

Different theories of AI policy. Anthropic's public posture has been pro-safety-regulation since at least 2024 — Dario Amodei has publicly supported some forms of mandatory safety testing for frontier models. OpenAI's posture has shifted over time but in 2026 leans toward opposing state-by-state rules while accepting some federal coordination. The super PAC spending operationalizes those policy differences in specific races, especially where state-level AI bills could create precedent or where individual legislators are visible AI-safety voices. The Bores race is the highest-leverage individual fight because of the RAISE Act's potential influence on other states.

Is AI super PAC spending influencing AI regulation already?

Yes, in the sense that state AI legislators across multiple states (California, New York, Colorado, Illinois) have publicly cited concerns about AI-industry political spending. Whether the spending actually changes outcomes is harder to assess — many of the races where AI PACs are active also have other major issues at play. The Guardian called the NYC race 'a key battleground in the AI civil war' and noted that AI-focused super PACs are now among the top political spenders in the 2026 midterms overall. The 2026 midterms are establishing precedent for AI-industry political activity that will scale into 2028.

Quick Answer

OpenAI vs Anthropic Super PACs: 2026 Midterms (Jun 2026)

Published: June 24, 2026

OpenAI vs Anthropic Super PACs in the 2026 Midterms

AI super PAC spending in the 2026 US midterms has now topped $185 million, according to reporting this week from Yahoo News, Gadget Review, Futurism, the Guardian, NPR, and the LA Times. OpenAI-backed groups have spent ~$23.5M; Anthropic-backed groups ~$16.6M. Roughly half the AI-PAC spending is concentrated on a single New York City congressional primary race involving Assemblyman Alex Bores, who sponsored New York’s RAISE Act AI safety legislation. Here’s what’s at stake.

Last verified: June 24, 2026.

TL;DR

Group	Reported 2026 midterm spend	Stance
OpenAI-backed super PAC	~$23.5M	Anti-state AI regulation, pro federal pre-emption
Anthropic-backed groups	~$16.6M	Pro safety regulation, especially RAISE Act-style provisions
Broader AI/tech-aligned PACs	~$145M (residual)	Mixed — mostly aligned with OpenAI posture
Total reported AI-aligned spending	~$185M+

Half the spending is concentrated on the NYC race involving Alex Bores. The fight is real, expensive, and likely to define the regulatory environment for AI through the late 2020s.

What’s actually happening

Per NPR (June 22, 2026), Yahoo News, Gadget Review, the Guardian, Futurism, and the LA Times (June 22-23, 2026):

1. Big numbers

Total AI-aligned super PAC spending in the 2026 midterms has now exceeded $185 million. That’s larger than many traditional industry-aligned PACs and approaches the scale of major single-industry political spending (energy, finance, pharma).

2. Bipartisan-style spending

AI super PACs are spending on both Republican and Democratic races. The cleavage isn’t partisan — it’s about AI policy stance (more vs less regulation).

3. The Bores / RAISE Act fight

The NYC congressional primary race involving New York State Assemblyman Alex Bores is drawing roughly half of the total AI-PAC spending. Bores sponsored New York’s RAISE Act, AI safety legislation that would impose obligations on frontier model developers.

An early anti-Bores ad — reported by NPR — argued laws like the RAISE Act would create “a chaotic patchwork of state rules that would crush innovation.” That language is the operational political argument of the OpenAI-aligned camp.

4. The spending mix

OpenAI-backed super PAC: ~$23.5M, broadly opposing state-level AI safety laws and supporting candidates who favor federal pre-emption
Anthropic-backed groups: ~$16.6M, broadly supporting candidates aligned with mandatory safety testing and the responsible-scaling-policy direction
Broader tech / AI-industry PACs: the residual ~$145M, mostly aligned with the OpenAI posture but not strictly under OpenAI’s umbrella

What the RAISE Act actually is

The RAISE Act (Responsible AI Safety and Education Act, as variously named in state legislatures) would impose on frontier model developers:

Mandatory third-party safety assessments before deploying models above a defined capability or compute threshold
Incident disclosure obligations for serious safety failures
Liability provisions for catastrophic harms attributable to deployed models
Pre-deployment testing requirements for certain high-stakes use cases

The exact text differs by state. California, New York, Colorado, and Illinois have all considered RAISE-like legislation in 2026. New York’s version (sponsored by Bores) is the closest to active passage.

OpenAI’s argument against state-by-state RAISE laws is that they create regulatory fragmentation, slow shipping, and put US labs at a competitive disadvantage vs Chinese frontier labs operating without similar constraints.

Anthropic’s argument for some form of RAISE-style legislation is that mandatory safety testing for the highest-capability models is necessary to avoid catastrophic harms and that state-level action is justified given federal inaction.

Both positions are sincerely held inside the respective labs. The political spending operationalizes them.

Why this fight is happening now

Three converging reasons:

1. Federal AI regulation is moving slowly

The Trump administration’s AI Executive Order (signed in early 2026) leaned light-touch federally. Congressional action on a comprehensive federal AI framework has stalled. That leaves the regulatory field open at the state level — where RAISE Act-style bills can pass without federal pre-emption.

2. The 2026 midterms set state-level precedent

The composition of state legislatures and US House members after 2026 will determine whether RAISE-style laws pass in California, New York, Illinois, Colorado, and several other large-AI states. Once those laws are on the books, they’re hard to undo and they influence national policy.

3. AI companies now have the capital

OpenAI’s last reported valuation puts it among the most valuable private companies in the world. Anthropic isn’t far behind. Both companies have meaningful political budgets relative to most policy fights. The infrastructure (super PACs, lobbyists, policy shops) is now in place to deploy at scale.

What the AI-industry spending signals

AI policy is now a real political category. Climate, healthcare, immigration, AI. The midterms are the first cycle where AI-policy spending is a top-tier political category.
Internal industry alignment is breaking. OpenAI and Anthropic spending against each other is qualitatively different from the AI industry presenting a unified front. The cleavage will deepen.
Sovereignty and pre-emption are the live questions. The federal-vs-state debate is the immediate political surface. Internationally, similar dynamics are unfolding around the EU AI Act and Asian sovereignty initiatives.
The Bores race is a leading indicator. If Bores wins (or his RAISE Act direction prevails), expect similar laws in more states. If the anti-RAISE camp wins, expect federal pre-emption pushes.

Risks to think about

Regulatory uncertainty as a cost driver. AI companies, AI infrastructure investors, and AI-adjacent businesses now have a multi-year political-risk premium baked into valuations. The June 23 AI stock sell-off includes some of this overhang.
Capture risk. Industry political spending raises legitimate concerns about regulatory capture. If AI safety laws are written by AI companies’ own lobbyists, the laws may be less stringent than the public interest would warrant.
Counter-mobilization. Civil society groups, academic AI safety researchers, and public-interest tech groups are starting to organize opposition. The political fight may broaden beyond OpenAI vs Anthropic.

What to watch from here

The Bores primary outcome — leading indicator for state-level RAISE momentum
California’s AI safety bill activity in the 2026 fall session — California is the most consequential state
OpenAI / Anthropic policy posture statements — formal lobbying disclosures will reveal more than super PAC spending alone
Federal AI bill movement post-midterms — whether a unified federal framework becomes possible after November 2026
EU AI Act implementation pressure on US labs — international regulatory dynamics influence US political calculus