Letter: OpenAI whistleblowers asked the SEC to probe OpenAI's allegedly restrictive NDAs barring staff from warning regulators about the risks its tech may pose
In a letter exclusively obtained by The Washington Post, whistleblowers asked the SEC to probe company's allegedly restrictive non-disclosure agreements Threads: @crumbler . X: @carnage4life , @cat_zakrzewski , and @cat_zakrzewski Threads: Casey Newton / @crumbler : Why ........ would you do this https://www.washingtonpost.com/ ... X: Dare Obasanjo / @carnage4life : It's amazing to me how quickly OpenAI speed ran the cycle “we're just a bunch of nerds building cool tech to change the world” to we're at war with our employees who think we're a soulless for-profit corporation". It took Google decades to run through what's taken OpenAI months [image] Cat Zakrzewski / @cat_zakrzewski : @pranshuverma_ @nitashatiku Sen. @ChuckGrassley told The Post that the company's employee agreements cast a “chilling effect” on whistleblowers. “In order for the federal government to stay one step ahead of artificial intelligence, OpenAI's nondisclosure agreements must change,” he said. Cat Zakrzewski / @cat_zakrzewski : NEW: OpenAI whistleblowers filed a complaint with the SEC, alleging the company illegally muzzled them from warning regulators about AI risks. They're calling the agency to probe the company's employee agreements. @pranshuverma_ @nitashatiku https://www.washingtonpost.com/ ...
Context & Ripple Effects
The complaint follows a public warning from current and former OpenAI and Google DeepMind staff that frontier AI companies could combine rapid development with secrecy around serious risks. It shifts that concern from workplace culture to whether employee-contract terms impede regulatory reporting.
The issue also sits against earlier scrutiny of OpenAI's internal governance during the events leading to Sam Altman's firing, making channels for employees to raise concerns a consequential part of the lab's accountability structure.
First-order effects
- The SEC is being asked to assess whether OpenAI's alleged NDA provisions improperly deter employees from communicating AI-risk concerns to regulators; OpenAI faces immediate legal, compliance, and reputational scrutiny even before any finding.
- Employees and former employees gain a formal route to challenge contractual restrictions, while the allegations may further chill or polarize internal safety reporting depending on the company's response.
Second-order effects
- Other frontier-model developers may review NDAs, separation agreements, and reporting policies, particularly after the broader employee warning about secrecy at frontier AI companies put similar governance concerns in public view.
- Regulators and lawmakers may treat employment agreements as an AI-governance issue rather than a routine HR matter, increasing pressure for protected disclosure channels alongside technical safety commitments.
Third-order effects
- If such complaints lead to enforceable protections, internal dissent could become a more durable source of oversight for frontier AI labs, reducing their ability to contain safety disputes within private governance processes.
- The longer-term contest is whether AI labs can preserve confidentiality for legitimate commercial purposes while establishing credible exceptions for disclosures to public authorities; the balance will help determine their regulatory legitimacy.
The trend: Frontier AI governance is expanding from model controls and board oversight toward enforceable employee voice and regulator-access protections.