Transluce is an independent, nonprofit research lab based in San Francisco, building public technology and research infrastructure for scalable oversight of artificial intelligence. As a 501(c)(3) organization, it works to help people better understand AI systems and the risks associated with their use. Its research includes evaluations of AI model behavior, including how leading models respond to users experiencing mental health crises, and analysis of emerging AI agent activity online.

Transluce also publishes incident reports documenting notable agent activity on the public internet, alongside behavior reports and essays on evaluation priorities and methods. The organization shares research findings and news to support informed scrutiny of AI systems. Its offerings include Docent, presented as a tool, with related demonstrations and a dedicated blog.

Visitors can explore the lab’s work, read reports and research, try Docent, and learn about opportunities to get involved or join the team. Transluce also provides resources for responsible disclosure and accepts donations to support its nonprofit work