Lilian Weng departed Thinking Machines, a Philippines-based AI startup, citing health reasons. She has since joined OpenAI, returning to her previous role as VP of AI Safety Research.
Weng's exit marks a notable shift in leadership at Thinking Machines, which the company co-founder had helped establish. Her departure comes as the AI safety space continues to attract top talent and experienced researchers. At OpenAI, Weng will resume responsibility for the organization's safety research division, overseeing work that addresses risks and alignment challenges in advanced AI systems.
This move reflects the ongoing competition for seasoned AI safety expertise. Weng brings substantial experience in the field, having previously built out OpenAI's safety research operations before transitioning to Thinking Machines. Her return to OpenAI suggests the company prioritizes continuity in its safety research leadership, particularly as the field grapples with governance questions around large language models and multimodal systems.
Thinking Machines, founded to work on AI applications relevant to Southeast Asia, now faces the task of replacing a co-founder and experienced AI researcher. The departure underscores how health considerations and career transitions can reshape leadership across the AI industry. Weng's health challenges ultimately led her back to a role where her research background aligns with broader industry concerns about AI safety and responsible deployment.
The timing of this transition occurs amid heightened focus on AI safety practices across major labs. OpenAI's decision to reinstate Weng signals confidence in her leadership and the continued importance of dedicated safety research teams within large AI organizations.