A $5M PAC is trying to punch above its weight against Big Tech’s war chest

There’s a new political action committee in town, and it’s trying to do something that sounds almost quaint in 2026: run on small donations from actual tech workers rather than corporate PAC money.

Guardrails launched with a clear pitch — it’s a populist movement funded by people who build AI systems, not the executives who profit from them. The group has raised about $5 million so far, which sounds like a lot until you realize that Big Tech spent over $100 million on lobbying last year alone.

That’s not a typo. $5 million versus $100 million. It’s like bringing a knife to a gunfight, but the founders insist they’re not naive.

“We know we’re outspent,” one of the organizers told me. “But we’re not trying to outspend them. We’re trying to out-organize them.”

The strategy hinges on the idea that politicians are more responsive to organized constituents than to checkbooks, especially when those constituents are actual voters in their districts. Guardrails is targeting primaries and key congressional races where AI regulation is actually a live issue, not just a talking point.

What I find interesting is who’s funding it. The donor list reads like a who’s who of mid-level engineers, researchers, and product managers from OpenAI, Google DeepMind, Anthropic, and Meta. These are people who’ve seen the sausage being made and are worried about what’s coming out of the grinder.

They’re pushing for three things: mandatory safety testing before deployment, transparency requirements for training data and model capabilities, and a federal AI agency with actual enforcement teeth. None of this is radical by global standards — the EU AI Act already mandates most of this — but in the US political landscape, it’s practically revolutionary.

Big Tech’s response has been predictable. The usual trade groups are running ads calling any regulation “innovation-killing” and warning that America will fall behind China. The same playbook we’ve seen for decades, just with AI replacing “net neutrality” and “encryption.”

But here’s where it gets interesting: the public is actually on Guardrails’ side. Polling consistently shows that 70%+ of Americans want stronger AI regulation, and that number cuts across party lines. The challenge is translating that diffuse public support into focused political pressure.

Guardrails is betting that $5 million, spent smartly on digital organizing and targeted ad buys in swing districts, can punch above its weight. They’re running a distributed canvassing program where AI workers volunteer to talk to their own representatives. It’s grassroots, but with a technical twist.

I’ve seen this approach before in other industries — climate activists tried it, net neutrality advocates tried it — and the results have been mixed at best. The problem is that Big Tech has a lot of tools beyond just money: revolving door hires, think tank funding, and a media apparatus that’s still largely deferential to Silicon Valley.

Still, I can’t help but root for them. The AI industry is moving so fast that even the people building it are scared of what they’re creating. If a bunch of engineers pooling their salaries can actually shift the political needle, that would be a genuinely new thing in American politics.

But let’s be real: $5 million is a rounding error for these companies. Google alone spent more than that on lobbying in the first quarter of this year. Guardrails is going to need a lot more than good intentions and small donations to win this fight.

What they do have is timing. The AI safety debate has moved from academic journals to kitchen tables, and the window for meaningful regulation is closing fast. If they can’t make this work now, with the public clearly on their side and the technology still young enough to steer, I’m not sure when they ever will.