OpenAI Staffers Warned Leadership of Security Lapses Years Before Recording Device Discovery

OpenAI employees raised internal concerns about lax security practices to top leaders and the board even before ChatGPT launched in late 2022.

Staffers argued that outsiders could easily access sensitive project details because code was stored in a dense single repository known as a monorepo.

Senior executive Thibault Sottiaux reassigned a quarter of production engineers in August to secure internal systems and improve security infrastructure.