Esben Kran

Esben Kran is an AI safety researcher, entrepreneur, and cognitive scientist focused on technical AI safety, model evaluation, and reducing risks from advanced AI systems. He is a co-founder and Co-Director of Apart Research and the founder of Seldon Lab, which supports work at the intersection of AI safety, security, and entrepreneurship.

Kran has a background in cognitive science, brain-computer interface research, and data science. His recent research includes DarkBench, a benchmark for identifying manipulative “dark patterns” in large language models, which was published at ICLR 2025 and received an Oral Spotlight. His broader work spans AI safety, model evaluation, and the development of safer advanced AI systems.