In Summer 2024, the former OxAI Safety and Governance team decided to spin out into an independent organisation: OAISI, the Oxford AI Safety Initiative.

We believe the AI safety community at Oxford would benefit from an organisation specifically trying to support research and education in AI safety here. There are many people who are highly capable of working on AI safety in Oxford who haven’t yet been introduced to the field or aren’t working on it.

Our purpose is to support the existing community, increasing the productivity, experience, and quality and quantity of conversations and projects about AI safety. Introduce new talent in Oxford to the field, and provide structure and support in building their skills. We believe AI safety is a technical problem as well as a governance problem, and as such, it requires both technical and policy talent.

Meet OAISI’s Executive Committee

  • James Lester

    President

    Hi, I'm James, and I'm at Balliol studying for an MPhil in Economics. Before this, I graduated from Cambridge (also economics), where I became concerned about extreme AI risks through the effective altruism community in 2022. I was an ERA:AI fellow in 2024, and have also worked in a think tank, economic consultancy and quant trading firm. My research has focused on incentive problems and "market failures" across AI governance and strategy.  Super happy to chat about OAISI, AI risk and economics in general, especially if you're from an EA, policy or less technical background!

    Book a Meeting
  • A young woman with long blonde hair and light skin standing outdoors in front of lush green leaves, smiling softly at the camera.

    Juliana Eberschlag

    Programme Manager

    Hi, I’m Juliana! I’m a Programme Manager at OAISI, where I work across operations, community building, and programme design to support Oxford’s growing AI safety ecosystem. Previously, I studied Philosophy at Queen’s University in Canada with a focus on philosophy of science, mind, AI, and logic - topics which remain reflective of my interests. I’m always keen to connect with others thinking about AI safety (especially from a high-level perspective), but I’d also be excited to chat all things music, personal psychology insights, or whatever else you’ve been pondering lately.

    Book a Meeting
  • Portrait of a man with long black hair, smiling, wearing a purple sweater and light blue collared shirt, standing outdoors with a blurred background of greenery and buildings.

    Matthew Farrugia-Roberts

    Committee Member

    Hi, I'm Matthew! I'm a DPhil in Computer Science at Magdalen. I came to Oxford to deepen my understanding of technology and its future and to broaden my understanding of society and its history, in order to help these things work better together. My current technical research is on understanding emergent agency in reinforcement learning. In the past, I have worked on developmental interpretability and singular learning theory, goal misgeneralisation, neural network geometry, and inverse reinforcement learning. I'm excited to meet students and researchers across Oxford who are studying and working on all topics relevant to the future of AI and society.

    Book a Meeting
  • Young man with dark hair and glasses smiling outdoors near yellow roses, field, and trees in the background.

    Louis Thomson

    Committee Member

    Hi, I'm Louis! I've recently graduated from the Master's in Computer Science and Philosophy at Oxford. My main interests include AI [control / safety / ethics], game theory, and cooperative AI. I've previously worked on the following projects (in order of recency): agentic monitoring for AI control; cooperation and alignment in multi-agent games; AI-Control games; and behavioural evaluations of deception in LLMs.

    I'm always up for talking about all things CS, philosophy (particularly AI-relevant topics but also continental philosophy), and other entirely-not-work-related stuff like music and games!

    Book a Meeting
  • A young man with curly, dark hair, wearing a light blue collared shirt, standing against a plain, light-colored wall.

    Rohan Selva-Radov

    Advisor

    Hey – I’m Rohan, a third-year undergraduate studying PPE at Merton!

    Last year I helped run OAISI, and I’m now still involved as an advisor. I previously worked on investing & strategy at Macroscopic Ventures, a non-profit grantmaker focused on AI safety & preventing dystopian futures. Outside of AI, I'm a big fan of bouldering and gardening, and I'd be excited to chat about any & all of the above!

    Book a Meeting