We are the central research and development team driving Faculty’s broader AI safety charter. We pioneer fundamental and applied technical AI safety research, producing rigorous scientific outputs—including publications, tooling, and technical reports—that advance the theory and practice of AI safety and directly shape the work of frontier AI labs, government agencies, and national security institutes.
Our research spans both black-box and white-box approaches to understanding and steering AI systems. We care deeply about interpretability, faithfulness, and principled mitigation strategies, with a particular focus on the mechanistic and behavioural study of transformer-based large language models and long-running agentic systems.
We operate at the cutting edge of the field, with current research threads including:
•
System Understanding: Interpretability, steering vectors, and the science of evaluations.
•
Control & Alignment: Goal drift, machine unlearning, and uncertainty calibration.
•
Societal Safeguards: Misinformation mitigation and robust deployment boundaries.
We collaborate closely with the Faculty’s wider AI safety team, a group with a well-established track record in capability evaluations and red-teaming for misuse risk across CBRN, cybersecurity, and international security. That work has been conducted for several leading frontier model developers and national safety institutes, and has been featured in OpenAI’s o1 system card and Meta Muse Spark’s Safety & Preparedness Report.
As a Senior Research Scientist at Faculty, you will join our high-impact R&D team to lead original technical research that advances our understanding of safe AI systems. Operating with high agency, you will balance scientific rigour with practical relevance to shape the future of safe AI deployment. Working closely with our broader AI safety practice, your scientific outputs—ranging from peer-reviewed publications to technical tooling—will directly inform frontier labs and government institutes, driving our entrepreneurial ambition to build a safer technological future.