Silicon Valley's Shadow Campaign: How AI Money Is Buying the 2026 Midterms Without Mentioning AI

The ads running in New York's 12th Congressional District look like any other political spot. One attacks a candidate for ties to ICE. Another hammers a rival on cost of living. A third warns about "right-wing billionaires" trying to buy a congressional seat. None of them mention artificial intelligence.

But artificial intelligence is the reason they exist.

Behind the ads are two rival super PAC networks, each backed by different factions of the AI industry, each spending millions to elect candidates who will shape how AI is regulated in America. It is the most expensive influence campaign the tech industry has ever waged in a midterm election — and it's being fought almost entirely in disguise.

The Two Armies

The AI industry's political spending in 2026 is organized around two competing umbrella organizations with fundamentally different visions for regulation.

The split is ideological but also commercial. Leading the Future's backers — including Andreessen Horowitz, one of Silicon Valley's most powerful venture capital firms — have billions invested in AI startups that could be constrained by aggressive regulation. Anthropic, which has positioned itself as the "safety-first" AI company, has a business interest in regulations that would disadvantage less cautious competitors.

The result is an intra-industry proxy war being fought with congressional candidates as pawns.

The Invisible Issue

What makes the AI spending campaign remarkable is not just its scale but its strategy. As NBC News reported, the super PACs' ads are about everything except AI. Immigration enforcement. Cost of living. Corruption. The bread-and-butter issues that move voters in competitive districts.

The logic is straightforward: voters don't go to the polls thinking about AI regulation. They go thinking about groceries, rent, and immigration. So the AI industry funds ads about groceries, rent, and immigration — and hopes the candidates who win will remember who paid for their victory when AI legislation comes to the floor.

It's a tactic borrowed from other industries. Pharmaceutical companies fund ads about veterans' healthcare. Oil companies fund ads about energy independence. The AI industry is simply the latest to discover that the most effective way to buy influence is to never mention what you're buying it for.

The New York Battleground

The earliest major battleground has been New York City's 12th Congressional District, where a Democratic primary has become a proxy war between the two AI factions.

Alex Bores, a former tech executive turned state legislator who has pushed for AI regulation, has become a primary target. Think Big — the Leading the Future affiliate that backs Democrats — has spent more than $1.5 million attacking Bores, including ads that hammer his former company's work with ICE. One digital ad declares: "It's black and white: Alex Bores' tech company works for ICE."

Jobs and Democracy PAC, the Public First affiliate, has countered with its own ads framing those attacks as cynical and profit-motivated: "Right-wing billionaires think they can buy this congressional seat, the same ones who bankroll hate, fund lies and prop up extremists."

A tech billionaire-backed super PAC network is reportedly prepared to spend up to $125 million across the 2026 cycle to undercut candidates who push for AI regulation — a figure that would make AI industry spending comparable to some of the largest traditional political donors in the country.

The Lobbying Machine

The campaign spending is just the visible tip of a much larger influence operation. According to Public Citizen, one in four federal lobbyists now works on AI issues — a staggering concentration of lobbying power around a single technology sector.

OpenAI and Anthropic both spent more on lobbying in 2025 than in any previous year, according to Forbes. The companies are pushing for favorable treatment from a Congress that is still struggling to understand the technology it's being asked to regulate. A Pew Research Center survey found that 69% of Americans believe government action on AI remains insufficient — suggesting that the industry's spending may be running against public sentiment.

Anthropic's Unusual Position

Anthropic occupies a particularly unusual position in the AI political landscape. The company gave up a Pentagon contract rather than loosen the safety guardrails on its AI model, Claude — a decision that cost it significant revenue but reinforced its brand as the "responsible" AI company. Simultaneously, it donated $20 million to Public First Action, a political group that opposes federal efforts to preempt state AI regulations.

The strategy is coherent if cynical: by supporting state-level regulation, Anthropic creates a patchwork of compliance requirements that larger, better-resourced companies (like itself) can navigate more easily than smaller competitors. Safety becomes both a genuine principle and a competitive moat.

What It Means for Democracy

The AI industry's entry into political spending represents something genuinely new. Unlike traditional industries that lobby to protect existing business models, AI companies are spending to shape the regulatory environment for a technology whose full implications are not yet understood — by the companies themselves, let alone by the legislators they're trying to influence.

The 2024 election saw $11.1 billion in total political ad spending. The AI industry's $83 million contribution was a rounding error. But the $125 million planned for 2026 — concentrated in specific competitive primaries where relatively small amounts of money can be decisive — represents a strategic escalation that could reshape which candidates make it to Congress.

The question is not whether AI money will influence the 2026 midterms. It already is. The question is whether voters will ever know that the ads about immigration and cost of living that fill their screens are really about something else entirely.

Sources: NBC News, The New York Times, OpenSecrets, TechCrunch, Forbes, Public Citizen, Reuters, Pew Research Center. All spending figures are from published reports and FEC filings as of March 2026.