AI companies don’t just build models — they build influence. Fresh numbers show how seriously Anthropic and OpenAI are taking Washington: in the second quarter of 2026, the two combined spent $3.17 million on federal lobbying. That’s 23 percent more than in the first quarter — and a record for both.
Anthropic outspends Nvidia
The split is the striking part. Anthropic alone put $1.97 million into lobbying — more than Nvidia, the chip giant half the industry leans on. OpenAI rose about 18 percent, landing at $1.2 million.
The lobbying targeted exactly the issues you’d expect when two AI heavyweights knock on the door of politics: cybersecurity, copyright, cloud computing and defense procurement. This isn’t about abstract principles — it’s about hard questions of who’s allowed to do what, with whose data, for which agency.
Why now
The timing is no accident. Both companies are heading toward an IPO and want a hand in shaping the regulatory guardrails before they’re set in stone. As models push deeper into agencies, the military and critical infrastructure, policy increasingly shapes the market. Lobbying stops being a side note and becomes part of product strategy.
My take
I find the number interesting from two angles. One, it shows how quickly research labs turn into political players — Anthropic, which likes to position itself as the cautious, safety-first company, is outspending Nvidia on lobbying. Two, it’s a sober signal: if you want to play in AI, you play in Washington too. Whether that’s good for anyone is a separate question. What’s clear is that the rules for the next AI decade are being written right now — and the big providers want to hold the pen.
Sources: