OpenAI Accused of Lax Security as Concerns Went Unheeded for Years

OpenAI found a hidden recording device in an executive meeting room, and several former employees said they were not surprised. According to Reuters, staff had questioned the company's loose security even before ChatGPT launched in late 2022, feeling management cared more about catching up with rivals like Google and Anthropic. Sources said any employee with code access could view technical details of many projects, so outsiders or even foreign governments could get sensitive information more easily. In October, the US Justice Department arrested a Saudi engineer working at OpenAI for hiding a recording device in a meeting room. OpenAI once kept much of its code in one huge shared system, and as AI agents can now search through such large amounts of data, the risks have grown. The company says it has spent heavily on security, with strict limits on sensitive projects. Experts note that adding protection may slow research and complicate the planned IPOs of OpenAI and Anthropic.