OpenAI dismisses three safety researchers after internal dispute

The company cited mishandling of sensitive information, but the researchers say they were punished for safety warnings.

By Felo News Desk · Published Fri, 09 Oct 2026 12:00:24 GMT

OpenAI announced on Friday that it had terminated three safety researchers, saying they breached internal policies on handling sensitive information. The three staff members – Mikita Balesni, Tomek Korbak and Jasmine Wang – contend they were fired for pressing the company to address AI safety risks.

What happened

According to OpenAI’s post on X, the three researchers were "parted ways" with the firm after an internal investigation found they violated clear policies on handling sensitive information. The company said the dismissals were unrelated to any safety concerns raised by the staff.

The trio had earlier sent a letter to OpenAI’s safety oversight groups outlining the circumstances of their dismissals and warning that the episode could chill a culture of open safety discussion.

What the reports add

Independent reported that the firings were first reported by The Wall Street Journal and described OpenAI’s statement as a defence against accusations that the firm puts corporate interests ahead of safety. Al Jazeera added that the researchers said they were let go because they prioritized safety over OpenAI’s near‑term corporate goals, and quoted Balesni’s X post asserting that belief.

Both outlets note that the researchers warned of a "chilling" effect on other staff. Independent highlighted that the researchers urged OpenAI to honour its promise to allow third‑party safety monitors inside the company.

What was said

OpenAI’s X post said the staff "violated clear policies on handling sensitive information" and that the company had "parted ways" with them after an investigation.

Balesni told Al Jazeera, "I believe we were fired for prioritizing safety over the near‑term interests of OpenAI as a corporation." Korbak expressed a similar view, saying he was terminated for raising concerns that OpenAI was "losing the ability to monitor what AI agents think, one of our best tools for catching when they misbehave".

How it came about

The dismissals follow a series of safety‑related events at OpenAI. In September the company launched the “dots” AI agent after scrapping the GPT‑6.1 rollout, a move that raised internal safety debates. Earlier in the month OpenAI released over 370 new mathematical results, prompting external criticism about the use of proprietary models for frontier research. In mid‑September the firm disclosed six new AI misalignment incidents and introduced a framework for tracking such cases, underscoring ongoing safety concerns within the organization. These developments provide context for the researchers’ warnings about a chilling effect on internal safety discourse.

Key facts

OpenAI terminated three safety researchers, citing policy breaches on handling sensitive information. (independent.co.uk)
The researchers – Mikita Balesni, Tomek Korbak and Jasmine Wang – say they were fired for raising AI safety concerns. (aljazeera.com)
OpenAI posted the termination statement on X, describing the separation as a "parted ways" after an investigation. (independent.co.uk)