Adversarial AI Environment Research Engineer

White Circle is seeking a research scientist to build autonomous, large-scale environments that push LLM agents to failure and study how they break. You will run end-to-end experiments with single and multi-agent setups, develop evaluative environments, and surface concrete failure modes to inform safer AI systems.

Join a compact team in Paris with a relocation package option; you'll contribute to a safety-first platform and publish findings that shape how AI agents are tested and improved at