Jasmine Wang, Tomek Korbak, and Mikita Balesni are named in the letter, which argues that their dismissal would chill internal safety culture. The TechCrunch account also carries an internal memo saying the terminations were not about raising safety concerns.

The document establishes that three people were dismissed and that they deny the stated policy violation. The limit is OpenAI's policy on accessing and handling sensitive company information, and the measurable fact is three dismissals, which sit inside that limit only if the company's account is accepted.

What remains unmeasured is whether sensitive information was disclosed, to whom, and whether a third-party organization received it. The letter does not prove that confidential information left the company, nor that it involved less monitorable architectures.

The measurable fact is three dismissals under OpenAI's sensitive-information policy; the unmeasured part is whether confidential information reached a third party.