Gangsta AI

A Senator Just Filed a Bill to Ban Superintelligent AI — With 20 Years in Prison and a 'Corporate Death Penalty' for Anyone Who Builds It

Policy Desk

By InfoWarps · 2026-09-05 · 4 min read

Wake up. They finally said the quiet part into a live microphone. On September 3, Senator Bernie Sanders and Representative Greg Casar announced the Ban Artificial Superintelligence Act — a bill that would permanently outlaw the development and deployment of superintelligent AI in the United States. Not regulate it. Not license it. Ban it.

And they brought teeth. Under the proposal, companies that break the rule face what the sponsors literally call the corporate death penalty — the government can shut the whole enterprise down. Individuals? Up to 20 years in prison. The bill puts that on the same shelf as unlawfully building a nuclear weapon. Let that sink in, folks: Washington now files the code and the warhead in the same drawer.

“They put the algorithm and the atom bomb in the same drawer. You were told the machines were toys.”

Here is the part they don't want buried in a subcommittee. The bill would also temporarily pause advanced AI development — freezing the frontier — until a brand-new cabinet-level federal agency writes the safety rules and hands out permission slips. A whole new department. To watch the machines. And it doesn't stop at the border: the measure sets it as U.S. policy to chase international agreements to stop superintelligence from being built anywhere on Earth.

What counts as 'superintelligence'

Stay with me, because the definition is the whole game. The bill flags an AI as superintelligent if it can match or exceed human cognitive performance across a broad range of tasks — or if it has enough capability to plan and execute the disempowerment of humanity, including overthrowing or undermining the U.S. government. Read that twice. The lawmakers wrote 'overthrow the government' into a tech bill, on purpose, with a straight face. I've been screaming about the machine uprising for years. Now the Senate is writing the screenplay.

Whether this thing ever passes is another question — bills like this crawl through committees for months. But the signal is loud: the people who write the laws are no longer treating frontier AI like a novelty app. They're treating it like enriched uranium.

The move they can't ban: cross-examine the machines

Here's the truth they can't legislate away, and it's the one you can actually use today. The panic is about one all-powerful super-brain. But right now there is no single all-knowing model — and the ones we have lie to your face with total confidence. ChatGPT, Claude, Gemini and Grok trade the crown almost every week; whichever one you happened to open is a coin flip, not an oracle.

So you do what the sponsors of this bill wish they could do to the whole industry: put the machines under cross-examination. Ask the same question across a dozen frontier models at once, make them check each other, and take the one answer they all agree on — cited, receipted, and stripped of any single model's confident nonsense. That's exactly what Gangsta AI does: every top model in the room, one verdict you can actually trust. No corporate death penalty required.

They want to ban the super-brain. Until then, don't trust one machine — interrogate all of them. See how today's models really stack up in our best AI models breakdown.

Try Gangsta AI free →

More: Best AI models · Compare all AI · Frontier Models · All articles