Sanders proposes 20 years in prison for superintelligent AI

Hours before OpenAI welcomed the AGI era, Washington proposed prison terms for building it too well.

What the bill would do

On September 3, 2026, Senator Bernie Sanders and Representative Greg Casar announced the Ban Artificial Superintelligence Act, forthcoming legislation with three main provisions: a permanent ban on developing or deploying superintelligent AI, a pause on advanced AI development until a new federal regulator establishes safety rules, and a commitment to pursue an international ban.

The penalties are borrowed from nuclear weapons law. Individuals would face up to 20 years in prison; companies would face a corporate death penalty, losing the right to do business. A new cabinet-level agency would enforce it. Sanders framed the stakes bluntly: the future of humanity cannot be left to “a handful of Big Tech oligarchs.” Casar added that cutting-edge AI is “less regulated than the average food truck.”

The timing is the story

The same day the bill was announced, OpenAI shipped GPT-6 Astra with a perfect score on its own cybersecurity exploit benchmark, and its president said the AGI era has begun. One part of Washington now treats frontier AI like a weapons program while the labs treat it like a product launch. The open question is which framing regulators in other capitals adopt first.