OpenAI Overhauls Security After Engineer Arrested for Espionage

OpenAI is restructuring its internal security infrastructure after the United States Department of Justice arrested an engineer, Abdulrahman Alfozan, for espionage. Alfozan, a Saudi national, allegedly installed a recording device in an executive conference room and used internal tools to track executive visits to Saudi Arabia.

The breach follows years of warnings from employees that internal security weaknesses could allow foreign governments to access sensitive technology. Staff cited concerns over a single monorepo providing broad access to company code, insufficient security reviews for research tools, and a corporate culture that prioritized competition with rivals like Google and Anthropic over security protocols.

In response, OpenAI has shifted its operational focus. Senior executive Thibault Sottiaux recently reassigned 25% of production engineers to security infrastructure, and the company appointed Dane Stuckey as Chief Information Security Officer in October 2024. While the company maintains that its most sensitive work remains in restricted repositories, it recently launched a new internal security initiative to address the reported vulnerabilities.