OpenAI released a 13-page policy paper proposing solutions for AI’s impact on American workers. The timing couldn’t have been worse. As The Verge AI reports, the paper dropped the same day The New Yorker published a damning 17,000-word investigation into Sam Altman’s pattern of saying one thing and doing another.
The proposals themselves are genuinely interesting: higher capital gains taxes on companies replacing workers with AI, a public wealth fund, a four-day workweek funded by “efficiency dividends,” and government programs to transition workers into “human-centered” roles. On paper, it reads like a progressive blueprint for managing AI’s labor disruption.
But Washington isn’t buying it. At least not yet.
📌 The Credibility Problem
The core issue isn’t what OpenAI proposed. It’s whether anyone believes they’ll follow through. Multiple sources told The Verge AI that the paper introduced valuable ideas into the policy conversation, but OpenAI’s track record makes the whole exercise feel performative.
Malo Bourgon, CEO of the Machine Intelligence Research Institute, captured the tension perfectly. He acknowledged that people on OpenAI’s team likely “care about the stuff” and “did good work.” But then he asked the real question: will those people end up disillusioned and leaving, like so many before them?
This isn’t hypothetical. OpenAI has a well-documented pattern:
- Altman publicly advocated for a federal AI oversight agency in 2023, then privately worked to suppress bills containing his own safety proposals
- A California legislative aide accused OpenAI of “increasingly cunning, deceptive behavior” to kill an AI safety bill it publicly supported
- In 2025, the company subpoenaed supporters of a California AI bill to intimidate them into silence
- Altman helped the Biden administration build AI safety standards, then convinced Trump to dismantle them
Nathan Calvin, general counsel at AI policy nonprofit Encode (and a recipient of one of those subpoenas), called OpenAI’s policy engagement “abysmal.” He believes the technical safety researchers who wrote the paper had good intentions, but remains deeply skeptical about what happens when principles meet lobbying.
📌 Why This Matters Right Now
OpenAI is trying to shape AI regulation while simultaneously undermining it. That’s not a contradiction the company can sustain forever. Washington has a long memory, and the people writing these rules are paying attention.
What stands out here is the growing gap between OpenAI’s public positioning and its private behavior. Every AI company lobbies. Every company tries to shape regulation in its favor. But few companies have built their entire brand on safety and responsibility while actively sabotaging safety legislation behind closed doors.
This credibility deficit has real consequences. When the company that dominates the AI conversation can’t be trusted to mean what it says, it poisons the well for the entire industry. Policymakers become more cynical. Good-faith proposals from other companies get met with suspicion. And meaningful regulation gets harder to build because nobody knows which corporate commitments are real.
📌 What AI Practitioners Should Watch
The substance of OpenAI’s proposals deserves serious discussion, regardless of who’s proposing them. A public wealth fund funded by AI productivity gains is a concept worth exploring. The four-day workweek idea, while vague, at least acknowledges that AI’s economic benefits need to be distributed, not just concentrated.
But the lesson for the broader AI industry is clear: policy credibility is earned through consistent behavior, not white papers. Companies that want a seat at the regulatory table need to align their lobbying with their public statements. The moment those diverge, everything becomes noise.
OpenAI declined to comment to The Verge AI. The full analysis, including detailed sourcing on OpenAI’s legislative history, is worth reading at the original source.