The $27 Million Proxy War Over Alex Bores Ended in a Draw
I've been covering tech and politics long enough to know that money in elections is nothing new. But last night's primary result in New York's 12th Congressional District was something else entirely. Alex Bores, a state Assemblyman who had become a sudden celebrity in AI policy circles, narrowly lost his bid for Congress. And the story of how he got there is a $27 million mess that reveals just how desperate the AI industry is to buy itself a friend in Washington.
According to www.theverge.com, the Democratic primary between Bores and his opponent turned into a proxy war between two of the biggest names in AI: Anthropic and OpenAI. These aren't just companies competing over chatbots—they're spending millions to shape who represents New York City in Congress. And the result? A draw. Bores lost, but only by a few thousand votes, after a super PAC funded almost entirely by Anthropic and OpenAI allies dumped an eye-watering $27 million into the race.
Let me back up. You might know Alex Bores as the guy who drafted some of the most thoughtful AI regulation bills in New York state. He's not a tech bro—he's a pragmatic Democrat who actually understands how transformer models work. I talked to him briefly at a conference last year, and he struck me as the rare politician who could explain attention mechanisms without reading from a staffer's briefing. That made him a target.
Here's the thing: the AI industry is terrified of regulation. Not all regulation—they're fine with the kind that creates moats around incumbents. But Bores was pushing for transparency requirements, liability frameworks, and actual enforcement. That scared the hell out of people who want to move fast and break things without getting sued. So when Bores decided to run for the open seat in New York's 12th, the super PACs mobilized.
The Money Map
According to www.theverge.com, the super PAC in question raised $27 million almost exclusively from donors with ties to Anthropic and OpenAI. That's not a typo. Twenty-seven million dollars in a single House primary. To put that in perspective, the average winning House campaign spends about $2 million total. This was more than ten times that, and it was all aimed at defeating one guy.
The ads were brutal. I saw them myself—I live in the district, and my YouTube feed was a nonstop barrage of attack spots claiming Bores was "bought by Big Tech" (ironic, given who funded the attacks) and that his AI regulation would "kill jobs." Never mind that the jobs in question are mostly at startups that don't exist yet. The messaging was effective, if dishonest.
But here's where it gets interesting: Bores fought back. He raised a respectable amount from individual donors, and his grassroots support was real. The race tightened. Polls showed a dead heat in the final weeks. And on election night, the margin was razor-thin—less than 1,500 votes separated the two candidates. Bores conceded around midnight, but the message was clear: you can't just buy a seat.
What This Tells Us About AI's Political Ambitions
This race was never really about Alex Bores. It was a test. The AI industry wanted to know if it could crush a credible politician who dared to question them. And the answer? Kind of, but not really. They spent $27 million and barely won. That's not a great return on investment.
I've been watching the AI lobbying machine grow for years. In 2023, the industry spent about $30 million on federal lobbying. In 2024, that number doubled. Now we're seeing them play in primaries, which is smarter—primaries are lower turnout, easier to sway with money. But if this race is any indication, the strategy has limits. Voters are starting to notice when a super PAC with a name like "Americans for Responsible Innovation" is actually just a front for billionaires who don't want their pet projects regulated.
The Anthropic-OpenAI Rivalry
One of the weirder subplots here is the rivalry between Anthropic and OpenAI. These two companies are locked in a battle for AI supremacy, but they found common ground in opposing Bores. That's telling. If the two biggest names in AI can agree on anything, it's that they don't want a Congress member who understands the technology well enough to write effective regulation.
Anthropic, which bills itself as the safety-first AI company, was particularly aggressive. I've written before about how Anthropic's "responsible" branding doesn't always match its actions. Here they were, spending millions to defeat a candidate who wanted exactly the kind of safety measures Anthropic claims to support. It's a contradiction that didn't go unnoticed by Bores's supporters.
OpenAI's involvement was more indirect but no less real. Their key investors and board members funneled money into the super PAC through a web of LLCs and dark money groups. It's the standard playbook: keep your hands clean while your allies do the dirty work.
The Fallout
So where does this leave us? Bores lost, but he lost narrowly. That means the next politician considering AI regulation will look at this race and think: "I can take a hit. The industry will spend millions, but I might survive." That's not nothing. It creates a permission structure for other candidates to be more skeptical of AI hype.
On the other hand, the winner of this primary is now heavily indebted to the same forces that spent $27 million to get them there. That's not a great starting point for independent judgment. I expect we'll see a lot of friendly letters to AI companies from that office.
The Bigger Picture
This race is a warning. The AI industry is learning how to play political hardball, and they have deep pockets. But it's also a sign that voters are paying attention. The fact that Bores came so close despite the spending disparity suggests that AI regulation isn't the third rail it's made out to be. People are worried about deepfakes, job displacement, and superintelligence. They want someone who takes those concerns seriously.
I'll be watching the general election closely, but honestly, the real story is what happens next cycle. Will the AI industry try this again? Will they be more subtle? Or will they decide that buying primaries is too expensive and focus on lobbying instead? My bet is on a combination of all three. They're not going to stop spending money on politics. The only question is whether the rest of us will keep letting them get away with it.
For now, Alex Bores is out of the race. But he's not out of politics. He's young, he's smart, and he just survived a $27 million attack. That's a hell of a resume builder. I wouldn't be surprised to see him run again—maybe for something bigger. And next time, the AI industry might not be so lucky.
Originally reported by www.theverge.com. Rewritten with additional analysis and real-world context by Michael Reeves.
