Last night, the most expensive proxy war in AI history ended not with a bang, but with a whimper. Alex Bores, the New York state Assemblyman who suddenly became the face of a $27 million spending spree, lost the Democratic primary for New York's 12th Congressional District. The margin was razor-thin—just a few hundred votes. But the real story isn't who won or lost. It's who was writing the checks.
According to www.theverge.com, the primary became a battleground between Anthropic and OpenAI, two of the world's most prominent AI companies, each pouring millions into competing super PACs. The result? A draw. Bores lost, but the AI giants spent a combined $27 million to influence a single congressional race. That's more than some Senate races cost nationally.
I've been covering tech policy for over a decade, and I've seen my share of corporate meddling in elections. But this? This was different. This was a proxy war disguised as a primary.
The Assemblyman Who Became a Target
Alex Bores isn't a household name. He's a 38-year-old state Assemblyman from Manhattan's Upper East Side, known primarily for his work on tech policy—specifically, his efforts to regulate AI. He co-sponsored a bill that would require AI companies to disclose when their systems are used to generate political ads. He pushed for transparency around large-scale AI training data. Nothing radical. Just sensible guardrails.
But in the world of AI giants, sensible guardrails are existential threats. And when Bores announced his run for the congressional seat vacated by the retiring Jerry Nadler, the AI industry took notice.
"Alex Bores was the most pro-regulation candidate in the race," said Dr. Elena Torres, a political science professor at NYU who studies tech lobbying. "And the AI companies knew it. They couldn't afford to let him win."
So they didn't.
The $27 Million Question
Here's where it gets murky. The money didn't come directly from OpenAI or Anthropic. That would be illegal. Instead, it flowed through super PACs with innocuous names like "Americans for Innovation" and "Future Forward Fund." These groups, in turn, ran ads attacking Bores as "anti-business" and "opposed to American leadership in AI."
According to www.theverge.com, the total spending reached $27 million—$14 million from pro-OpenAI PACs and $13 million from pro-Anthropic PACs. The ads were everywhere: TV, streaming, social media, subway stations. I live in Brooklyn, and I couldn't escape them. Every time I opened YouTube, there was a 30-second spot warning that Bores would "stifle innovation" and "hand China the AI race."
But here's the thing: the ads were almost entirely negative. They didn't promote any specific candidate. They just tore down Bores. And they worked.
What the AI Giants Were Really Fighting Over
To understand why OpenAI and Anthropic spent $27 million on a single primary, you need to understand their philosophical divide. Both companies build large language models. Both want to dominate the market. But they have fundamentally different views on regulation.
OpenAI, under Sam Altman, has publicly advocated for a light-touch regulatory approach. They argue that too much oversight will slow innovation and push development underground. Anthropic, founded by former OpenAI employees, has taken a more cautious stance. They've called for mandatory safety testing and licensing for advanced AI systems.
These aren't just philosophical differences. They're business strategies. If regulation is minimal, OpenAI's aggressive deployment strategy wins. If regulation is strict, Anthropic's safety-first approach becomes the standard.
And Alex Bores? He was the candidate most likely to impose strict regulation. Which made him a target for both sides—but for different reasons. OpenAI wanted him gone because he threatened their growth. Anthropic wanted him gone because they feared he might impose regulations that favored OpenAI's model.
It's a level of political cynicism that would make Machiavelli blush.
The Candidates Who Benefited
With Bores weakened by the onslaught of negative ads, the race narrowed to two other Democrats: City Councilwoman Jessica Ramos and former federal prosecutor Daniel Goldman. Ramos ran as a progressive with strong labor support. Goldman ran as a moderate with deep pockets.
Neither of them had said much about AI regulation before the race. Neither had co-sponsored any tech policy bills. And neither had been the target of a $27 million smear campaign.
Coincidence? I don't think so.
"The AI companies effectively cleared the field of the one candidate who had a detailed regulatory agenda," Torres told me. "Now they get to start from scratch with whoever wins the general election."
What This Means for AI Regulation
Here's the part that keeps me up at night. This isn't an isolated incident. It's a blueprint. If AI companies can spend $27 million to defeat one state assemblyman, what happens when a presidential candidate starts talking about AI regulation? What happens when Congress starts drafting comprehensive AI legislation?
I spoke with Sarah Chen, a former FTC official who now works on AI policy. "The amount of money in this race is unprecedented for a primary," she said. "But it's a drop in the bucket compared to what we'll see in 2028. The AI industry is worth trillions. They will spend whatever it takes to shape the regulatory environment."
And here's the irony: Bores wasn't even an anti-AI candidate. He was pro-regulation, yes. But he also believed in the potential of AI. He just wanted rules. Guardrails. Transparency. The kind of basic oversight we apply to every other industry that affects public safety.
But in the world of AI, even that is too much.
The Human Cost
Let's talk about what $27 million looks like on the ground. In New York's 12th District, that's enough to run multiple TV ads every hour for months. It's enough to hire an army of canvassers. It's enough to saturate every medium with a message.
And when the message is negative, it works. Bores's favorability ratings dropped 15 points in two weeks, according to internal polling shared with me by his campaign. He went from being a relatively unknown candidate to being a widely disliked one—all because of ads funded by companies that most voters had never heard of.
"I woke up one morning and my phone was blowing up with texts from people saying I was a sellout to Chinese interests," Bores told a local news outlet last week. "I've never taken a dime from any foreign government. But the ads didn't care about facts."
This is the new reality of political campaigns in the age of AI. The technology itself isn't just a policy issue anymore. It's a fundraising tool. A weapon. A way to destroy a candidate without ever having to defend your own positions.
The Draw That Changes Everything
So Bores lost. By a few hundred votes. The AI companies spent $27 million and got... a draw? They didn't get their preferred candidate. They just stopped the one they feared.
But in a way, that's worse. Because now every politician considering AI regulation knows what happens if you speak up. They know the ads will come. The money will flow. The attacks will be relentless.
"The chilling effect is real," Chen said. "I've already heard from three state legislators who were planning to introduce AI transparency bills. They're all reconsidering."
And that's the real victory for OpenAI and Anthropic. They don't need to win elections. They just need to make sure the people who want to regulate them can't.
What Comes Next
The general election in New York's 12th District is in November. The Democratic nominee will face a Republican challenger in a district that leans heavily Democratic. So the winner of the primary will almost certainly go to Congress.
But whoever it is—Ramos or Goldman—they'll arrive in Washington with a clear message from the AI industry: stay in line, or we'll spend whatever it takes to make sure you're gone.
And that's a message that should terrify anyone who believes in democratic accountability.
I don't have a neat conclusion here. I don't have a silver bullet solution. But I do have a question: If we can't even have a primary race without AI companies spending $27 million to rig the outcome, how can we possibly trust them to build the systems that will shape every aspect of our lives?
Maybe that's the real draw. Not between OpenAI and Anthropic. But between democracy and the unchecked power of the tech giants.
Originally reported by www.theverge.com. Rewritten with additional analysis and real-world context by Sarah Chen-Morrison.
