Nine candidates that AI-focused super PACs supported with general-election spending had been declared winners by AP as of 02:33Z on 4 November, by our count from Federal Election Commission filings and AP’s race calls. None of the candidates those super PACs spent against in the general election was in a race AP had called by then.
The nine called races account for $1.81 million of the $19.85 million that nine AI super PACs reported in general-election independent expenditures across 35 races, or 9.1%, by our arithmetic from the FEC’s 24- and 48-hour reports. AP lists eight of the nine winners as incumbents; the other, Andy Barr, ran for the Kentucky Senate seat AP lists as held by Mitch McConnell.
The key insight: By our count from FEC filings and AP race calls, as of 02:33Z AP had called 9 of the 35 races in which nine AI super PACs reported general-election spending, and the candidates those PACs supported won all 9. Those races hold $1.81 million of $19.85 million.
AP has called Pennsylvania’s 13th District U.S. House race for John Joyce (GOP), and American Mission reported $240,452 supporting Joyce. AP has called the West Virginia U.S. Senate race for Shelley Capito (GOP), and American Mission reported $210,193 supporting Capito. AP has called Georgia’s 7th District U.S. House race for Rich McCormick (GOP), and American Mission reported $207,640 supporting McCormick. AP has called Florida’s 15th District U.S.
House race for Laurel Lee (GOP), and American Mission reported $207,329 supporting Lee. AP has called the Kentucky U.S. Senate race for Andy Barr (GOP), and American Mission reported $207,034 supporting Barr. AP has called Georgia’s 3rd District U.S. House race for Brian Jack (GOP), and American Mission reported $203,509 supporting Jack. AP has called Florida’s 8th District U.S. House race for Mike Haridopolos (GOP), and American Mission reported $194,627 supporting Haridopolos.
AP has called Ohio’s 13th District U.S. House race for Emilia Sykes (Dem), and Think Big reported $332,071 supporting Sykes. AP has called New York’s 15th District U.S. House race for Ritchie Torres (Dem), and Think Big reported $5,000 supporting Torres.
In New York’s 12th District, where our earlier piece counted $27.26 million in AI super PAC spending in the Democratic primary, AP has called the general election for Micah Lasher (Dem). Alexander Bores, a candidate in that primary, is not among the candidates AP lists for 3 November.
General-election independent expenditures supporting each candidate AP had declared a winner as of 02:33Z on 4 November 2026, from the FEC’s 24- and 48-hour reports (our sums). Race calls: AP. Chart: FourWeekMBA.
Where the Money Is Still Uncalled
Most of the general-election money is in races AP had not called at 02:33Z. The five largest amounts by candidate: Building the Future ($2.51 million) and American Mission ($1.60 million) reported a combined $4.11 million supporting Darline Graham (GOP) in the South Carolina U.S. Senate race; Tomorrow Together ($1.52 million) and Guardrails Alliance ($115,000) reported a combined $1.64 million opposing Jon Husted (GOP) in the Ohio U.S.
Senate race (unexpired term); Tomorrow Together reported $1.01 million opposing Roger Marshall (GOP) in the Kansas U.S. Senate race; Tomorrow Together reported $965,149 opposing Juan Ciscomani (GOP) in Arizona’s 6th District U.S. House race; Tomorrow Together reported $896,250 opposing Tom Barrett (GOP) in Michigan’s 7th District U.S. House race.
All nine called races drew spending from two Leading the Future-network PACs, American Mission (7) and Think Big (2). Four Public First-network PACs reported general-election spending, $10.54 million in all by our arithmetic; none of the races where they spent had been called.
Data Centers on the Ballot
AP had not called Ohio’s governor race at 02:33Z. Ohio holds 21 of the 41 local data center measures on 3 November ballots. Those measures are reported by county boards of elections, and we report their official returns in the results box on our Ohio piece. The full list is in our guide to the 41 measures.
Business Pill · Concentration
An evergreen Business Engineer explainer on concentration: when a few big pieces make most of a total. It teaches the general idea only and says nothing about any committee or candidate in this story.
How We Counted
We matched every general-election line in the FEC’s bulk file of 24- and 48-hour independent-expenditure reports, received through 30 October, for nine AI-focused super PACs to AP’s race list. We dropped superseded amendments and summed by committee, candidate and support or oppose. The grouping into two networks follows our earlier piece.
Results are AP race calls only. AP’s results page says: “AP’s Decision Team declares winners when certain the trailing candidates no longer have a path to victory.” Races AP had not called are described as uncalled.
The Structural Read
As we read the FEC lines, the general-election money is concentrated: the five largest candidate totals add up to $8.62 million of the $19.85 million, 43% by our arithmetic.
As we read AP’s data, the first calls came mostly in seats with an incumbent on the ballot: AP lists 8 of the 9 winners as incumbents.
AP, on how it calls races
“AP’s Decision Team declares winners when certain the trailing candidates no longer have a path to victory.”
Three Implications
FOR THE LEADING THE FUTURE NETWORK Of the 35 races where AI super PACs reported general-election spending, AP had called 9 by 02:33Z, and American Mission or Think Big reported spending in each of them, by our count.
FOR THE PUBLIC FIRST NETWORK None of the races where Public First-network PACs reported general-election spending had been called by 02:33Z, according to AP.
FOR DATA CENTER MEASURES Local measures are reported by county election offices; our results boxes quote official returns only, from those offices.
The framework defines its top layer, Governance, as “The rules that determine whether a model can be released, deployed, sold, or used.”
As we read it, Congress is one of the bodies that writes such rules. The framework is the owner’s argument and does not discuss the midterms or any candidate.
The called races are an early subset: 9 of the 35 races where these PACs reported general-election spending. The filings show where the money went, not what effect it had, and we do not attribute any result to it.
Spending that appears only in periodic FEC reports, and not in 24- or 48-hour reports, is not in these sums. Calls made after 02:33Z are not reflected here.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
By 02:33Z, AP had called 9 of the 35 races in which AI super PACs reported general-election spending, and the candidates they supported won all 9. The races holding most of the money were uncalled.
A note on sourcing. Race results are AP race calls only, read from AP’s results data at the time stated; races AP had not called are described as uncalled. Spending figures are our sums of the FEC’s 24- and 48-hour independent-expenditure reports for nine AI-focused super PACs, general election only. Shares and counts are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
