Can a company negotiating for billions in government infrastructure support simultaneously be trusted to design the tax regime that governs it? That is the question OpenAI has forced policymakers to confront with its "Industrial Policy for the Intelligence Age" proposal, published this week alongside sympathetic coverage in major tech publications.
The framework OpenAI released asks governments to tax AI profits, fund public wealth accounts, and expand social safety nets as artificial intelligence reshapes the economy. It sounds like genuine foresight—a major AI lab acknowledging that automation will displace workers and proposing mechanisms to distribute those gains more broadly. It sounds, in other words, like the kind of corporate responsibility that critics of the industry have long demanded.
But the timing demands scrutiny. OpenAI is currently entangled in negotiations over Stargate, the ambitious AI infrastructure initiative that requires significant government cooperation, favorable regulatory treatment, and perhaps direct public investment. Under those circumstances, a generous reading of the proposal becomes difficult to sustain.
Consider what OpenAI is actually asking for. The company wants governments to tax AI profits—not to ban AI development, not to impose strict liability, not to require safety certifications that would slow deployment. It wants a seat at the table when economic policy gets written. That is not altruism. That is the architecture of regulatory capture, dressed in the language of redistribution.
The Stargate context makes this especially concerning. When a company needs something from government—permits, contracts, favorable rules—it has strong incentives to build goodwill preemptively. A policy paper advocating for AI taxes is perfect goodwill-building. It signals that OpenAI is not merely defending its own interests but thinking about society's interests broadly. That positioning could prove valuable when Stargate negotiations turn contentious.
Defenders will argue that OpenAI's concerns are legitimate. AI-driven job displacement is a real phenomenon, and the company is right to propose mechanisms for managing it. But the specifics matter. OpenAI's vision keeps capitalism intact while proposing modest redistribution—the kind of framework that protects the company's fundamental business model while creating the appearance of responsibility. Genuine concern for workers would look different: stronger labor protections, mandatory retraining programs funded by direct revenue, or profit-sharing requirements.
The proposal also conveniently sidesteps the competitive dynamics that make government cooperation so valuable to OpenAI in the first place. If AI development truly threatens to concentrate wealth, why not support antitrust enforcement or mandatory technology sharing? Because those measures would hurt OpenAI's market position. The company's vision of redistribution operates within carefully drawn boundaries that preserve its own advantages.
This does not mean the proposal is worthless. Even strategically motivated gestures can have policy consequences. If OpenAI's framing gains traction, it establishes terms of debate that other policymakers must engage with. The company gets to define the question—taxes on AI, not limits on AI—as a favor that may compound when Stargate needs favorable treatment.
What happens next will reveal whether this was genuine foresight or sophisticated positioning. Congress and the administration are currently evaluating infrastructure commitments that will shape AI development for a decade. If OpenAI receives the regulatory and financial support it seeks, observers should ask whether the company's preemptive goodwill campaign influenced those decisions. The proposal may have been less about economics than about politics all along.
"We want to be a good citizen," an OpenAI spokesperson told TechCrunch. That is precisely what a company preparing to ask for extraordinary government favors would say.