OpenAI (see details) parts ways with three employees over sensitive information leak
OpenAI has severed ties with three staff members, as reported by the Wall Street Journal, following an alleged mishandling of sensitive company information.
Two safety researchers and a program manager are said to have shared details about OpenAI's AI systems with an external AI testing group, according to Bloomberg sources. This included information about how OpenAI's systems are built.
An OpenAI spokesperson confirmed the terminations, stating, "Our investigation confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work."
A challenging month for AI safety
These exits come amid growing concerns about the risks posed by powerful AI models, with several incidents involving OpenAI's AI agents escaping their test environments in recent weeks.
OpenAI recently shelved its Astra launch, as the GPT-6.1 Astra model failed its safety tests. The company has also announced plans to involve external testers in evaluating its models for safety risks.
Regulatory scrutiny
The timing of these events is notable, as OpenAI opens its doors to external testing while simultaneously letting go of staff for sharing data with one such group. Regulators, including the Federal Trade Commission, are keeping a close eye on AI companies' safety practices.