The New Yorker Drops a Sam Altman Bombshell, and OpenAI’s Superintelligence Pitch Looks Awkward

OpenAI had a weird day on Monday.

First, the company published a policy paper about how to handle superintelligence if it ever arrives — full of noble language about keeping people first, staying transparent, and monitoring for extreme risks like AI evading human control or governments using it to undermine democracy. They even warned that without proper safeguards, “people will be harmed.” It reads like a draft from an ethics committee that actually means it.

Then The New Yorker dropped a massive investigation into whether Sam Altman can be trusted to follow through on any of those promises. The headline sums it up: “The problem is Sam Altman.”

Reading both side by side is almost surreal. One document is all forward-looking optimism, proposing policies for a future where superintelligence means “a higher quality of life for all.” The other is a deep, uncomfortable look at the person who would be steering that ship.

I’ve been watching OpenAI’s governance circus for years, and this feels different. The New Yorker piece isn’t rehashing the boardroom drama from 2023. It’s digging into a pattern of behavior — trust issues that predate the firing and rehiring. Sources inside the company apparently don’t believe Altman means what he says, at least not consistently.

That’s a problem. You can’t credibly claim to be the responsible steward of superintelligence if your own employees don’t trust you to tell the truth about a product roadmap.

OpenAI’s policy paper tries to position the company as a mature, sober actor in the AI safety conversation. But investigations like this undercut that narrative. The timing feels either tone-deaf or desperately hopeful — like they thought they could control the news cycle by releasing something aspirational on the same day a major magazine was about to publish a hit piece.

To be fair, the policy recommendations themselves are reasonable. They talk about independent oversight, mandatory safety testing, and international coordination. If a different CEO were delivering this message, I might feel reassured. But Altman has a track record now, and trust is a currency you can’t fake.

What bothers me most is the gap between the rhetoric and the reality. OpenAI talks about transparency while operating behind closed doors. It advocates for democratic input while concentrating power in a few hands. The New Yorker piece just confirms what a lot of us suspected: the tension isn’t external. It’s inside the building.

Will any of this matter? Probably not in the short term. OpenAI is still the frontrunner, still raising billions, still signing deals. But the credibility gap is widening. If superintelligence really is around the corner, we might want the people building it to at least trust each other.