Archive · 8 Oct 2026
Former OpenAI Researchers Urge Independent Audits After Dismissals
Three former OpenAI researchers want independent experts to help check the company’s AI systems.
They say more powerful AI may become harder to understand and monitor.
They sent their recommendations in a letter to OpenAI’s board members and safety committees.
The researchers were recently dismissed after an investigation into how they handled confidential information.
OpenAI says the dismissals were about alleged rule breaches, not because they raised safety concerns.
The company says it agrees that outside evaluations can help and that monitoring AI is a priority.
OpenAI has not explained exactly what information was involved.
The disagreement has renewed discussion about how AI companies can keep their systems safe and open to scrutiny.
Three former OpenAI researchers urged the company to expand independent oversight of its AI systems.
Their letter to OpenAI board members and safety committees called for outside audits and an open safety research environment.
The researchers warned that increasingly capable models could become harder to understand and monitor.
OpenAI said the dismissals followed alleged breaches of rules on handling sensitive research information, not the raising of safety concerns.
OpenAI said it agreed with recommendations on monitoring and external assessment, while the details of the confidential material remain undisclosed.
- Who
- Three former OpenAI researchers; OpenAI
- What
- The researchers called for independent audits and stronger external oversight of advanced AI systems.
- When
- They sent the letter on Wednesday, shortly after their dismissals; the article gives no calendar date.
- Where
- The letter was sent to OpenAI board members and safety committees.
- Why
- They say increasingly capable AI systems may become harder to understand and monitor, and that outside scrutiny can help assess safety.
Former researchers
OpenAI
Need for independent oversight
Former researchers
The former researchers called for independent safety audits and greater external scrutiny of AI systems.
OpenAI
OpenAI said it agreed with recommendations for outside assessment and described monitoring as a major priority.
Reason for dismissals
Former researchers
The former researchers argued that the dismissals could discourage remaining staff from raising concerns.
OpenAI
OpenAI said the terminations concerned alleged improper handling of sensitive research material, not employees raising safety concerns.
- Researchers
- The Wall Street Journal reported their names as Jasmine Wang, Tomek Korbak and Mikita Balesni; the article says their identities were not independently confirmed by all reports.
- Letter recipients
- OpenAI board members and safety committees
- Key request
- Cooperation with independent safety auditors and support for an open safety research environment
- Dismissal explanation
- OpenAI said the terminations followed alleged breaches of company rules for handling sensitive research material.
- OpenAI's stated position
- The company said it strongly agreed with recommendations on AI monitoring and outside assessment.
- Unresolved detail
- OpenAI has not publicly specified what confidential information was involved or identified the outside organisation.