The company that convinced the world ChatGPT is infrastructure is now writing the playbook for how politicians should campaign on it.

The Summary

The Signal

This is lobbying dressed up as strategic advice. OpenAI wants candidates to make AI central to their campaigns, which makes sense when your business model depends on favorable regulation and endless access to power and data. What's remarkable is the audacity of the framing.

Lehane's memo positions any resistance to AI infrastructure buildout as a career-ending mistake on par with voting for the Iraq War. That's not analysis. That's threat construction. OpenAI needs massive new data centers to train increasingly large models. Local communities from Virginia to Oregon are fighting these facilities over water use, energy consumption, and grid strain. So Lehane is pre-framing opposition as geopolitical weakness, telling Democrats they'll hand China the future if they listen to constituents worried about their electric bills.

"If the US loses its AI lead to China, this could well be the 2028 version of the 2002 vote to go to war in Iraq."

The Iraq War comparison works both ways, though. That vote was driven by manufactured urgency, claims about existential threats, and pressure to prove toughness. Two decades later, most politicians who supported it call it a mistake. Lehane is betting candidates won't notice the parallel cuts against him.

The memo also urges a "new social contract" that keeps humans in control of consequential decisions. This is the diplomatic version of "we're building systems that will make consequential decisions, please don't regulate us too hard." The tension is built into the pitch:

  • Trust AI companies to build transformative systems
  • But also trust politicians to keep humans in charge
  • While building infrastructure those same politicians can't oppose without career risk

What makes this memo significant isn't that a tech company is lobbying. It's that OpenAI is shaping the governing philosophy itself, not just asking for favorable treatment within an existing framework. Lehane isn't requesting tax breaks or streamlined permits. He's telling future presidents how to think about the technology before they're even running.

The Implication

Watch how 2028 candidates talk about data centers. If they start using national security framing to justify local infrastructure fights, you'll know whose memo they read. The real test is whether any candidate pushes back on the premise that AI infrastructure is non-negotiable, or whether both parties accept OpenAI's framing that questioning the buildout is tantamount to surrendering technological leadership.

For anyone building in this space, note the strategy. OpenAI isn't waiting for regulation to happen to them. They're writing the narrative two years before the election. The company that defined "AI safety" as a brand position is now defining what responsible AI governance means in a campaign context. That's not lobbying. That's market making at the policy level.

Sources

TechCrunch AI | Business Insider Tech