OpenAI fires employees for sharing sensitive data with outside AI evaluation group

OpenAI has dismissed several workers following an internal investigation that found they had shared sensitive company information with an external AI safety evaluation group, BBC News reported on October 2, 2026. The company confirmed the terminations but declined to name the employees or the outside organisation involved.

What information was shared

According to reports, the dismissed employees had passed data to an outside group that conducts independent evaluations of AI systems. The nature of the data has not been fully disclosed, but it was described as sensitive enough to warrant termination. OpenAI said the employees violated the company's data-handling policies.

Implications for AI safety

The case raises complex questions about transparency in AI development. Critics argue that independent external evaluation of AI systems is essential to ensure safety, and that restricting such access could undermine public oversight. OpenAI, however, maintains that such information must be handled through formal, authorised channels.

The firings come at a time when the AI industry faces growing scrutiny over its practices. OpenAI has positioned itself as a leader in AI safety, and the incident is likely to generate further debate about how powerful AI labs balance openness with security.