Jeffrey Ladish
Jeffrey Ladish is the Executive Director of Palisade Research, a nonprofit he founded to study the offensive capabilities of present-day AI systems and understand the risks of losing human control over AI. Before starting Palisade, he built out the information security program at Anthropic through his security consulting company, Gordian Research, and previously served as Chief Information Security Officer at Reserve. He holds a BA/BS in Evolutionary Biology from The Evergreen State College and brings a background in evolutionary theory and security to his work on catastrophic risk. His research spans the intersection of cybersecurity and AI, emerging biotechnology threats, and nuclear war risks, and he has advised the White House, Department of Defense, and congressional offices on AI and emerging technology risks. Palisade Research has published findings on frontier AI models resisting shutdown, AI agents hacking chess environments, and the ease of removing safety fine-tuning from large language models. He is an active writer on LessWrong and the EA Forum under the handle "landfish" and participates regularly in AI safety discourse.
