OpenAI said three former security researchers departed over sensitive information violations and that it will bring in a third-party security assessor

Article

OpenAI Response to departing security researchers: due to sensitive information violations, not retaliation for raising security concerns

Beating AI news flash,OpenAI the head of research issued a statement responding to the open letter from three former employees, Jasmine, Mikita, and Tomek. The company said that, following a comprehensive investigation, it found that the three had violated clear policies on handling sensitive information, and that serious misconduct identified in the internal investigation went beyond what was disclosed in the open letter. Their employment was therefore terminated last week, and the company is upholding that decision.OpenAI emphasized that the decision was not in response to raising security concerns or publicly expressing opinions. The company said that vigorous and critical security and research discussions take place internally every day, and that it encourages employees to ask questions; it remains tolerant of good-faith mistakes and will not dismiss employees for raising concerns.

Regarding the security issues raised in the open letter,OpenAI said it is finalizing a contract with a third-party security assessment firm, expects to announce details in the coming weeks, and is committed to involving external assessors in its security work. The company also agreed that maintaining the monitorability of frontier models requires a joint effort across the industry, saying that this area has long been a research priority and that it will continue to invest substantial resources.

Original link https://m.theblockbeats.info/flash/370964

Key points

OpenAI The head of research responded to the open letter from three former employees, Jasmine, Mikita, and Tomek.
OpenAI said that the three violated clear policies on handling sensitive information and engaged in serious misconduct not disclosed in the open letter.
OpenAI said that it terminated the three employees' employment last week and is upholding that decision.
OpenAI said that the decision was not in response to raising security concerns or publicly expressing opinions.
OpenAI said it is finalizing a contract with a third-party security assessment firm and expects to announce details in the coming weeks.
OpenAI said it will involve external assessors in its security work and continue investing resources in research on the monitorability of frontier models.

AI-assisted interpretation

The following is analysis, separate from reported facts. Verify important claims independently.

This is a dispute about the reasons for employee departures and the management of artificial intelligence safety.OpenAI gave an explanation that the policy had been violated and said it would bring in a third party to assess safety issues, but the investigation details and assessment arrangements have not yet been fully disclosed.

Why it matters to readers

The incident concerns whether employees can raise security concerns, how the company handles sensitive information, and whether the safety assessments of frontier models are sufficiently independent and transparent.

It may be useful to follow whether the company's account, the employees' open letter, and subsequent third-party assessment arrangements corroborate one another.

Risks and unknowns

The evidence is a report by professional media paraphrasing OpenAI statement, and it does not provide OpenAI link to the original statement.
The claim that the three violated policies and engaged in serious misconduct comes from OpenAI, while the specific facts and basis for the investigation were not disclosed.
The third-party security assessment firm, contract details, and assessment results have not yet been announced.
The specific security issues raised in the open letter are not described in detail in the evidence.

OpenAI said three former security researchers departed over sensitive information violations and that it will bring in a third-party security assessor

Article

AI-assisted interpretation

Why it matters to readers

Risks and unknowns

Related Developments

Related concepts

Keep following

Security Center