Three former OpenAI safety researchers — Jasmine Wang, Mikita Balesni, and Tomek Korbak — have published an open letter disputing their dismissals, claiming the public nature of the firings created a chilling effect on internal dissent. They allege their terminations followed alleged information sharing with an external AI safety organization, which OpenAI cited as a policy violation involving sensitive data access. The researchers maintain their actions aligned with company mission and historical norms of open collaboration with external safety groups.
They warn the firings signal a shift in internal culture, leaving employees uncertain about acceptable conduct, especially regarding external safety engagement. The letter notes they were not the source of a recent The Information article on less monitorable AI architectures and emphasize their communications were coordinated with leadership. They urge OpenAI to uphold commitments to third-party safety audits, preserve model monitorability, and sustain an open dialogue with the broader safety ecosystem.
Jasmine Wang added on X that leadership appeared to agree with their letter, but expressed concern that speaking up or working with outside groups may now carry retaliation risks, warning others could face similar dismissals without explanation.
Source: Engadget · Summarized by HeadlinesBriefing