Ariel Gil - Professional Headshot

Ariel Gil

AI Safety Researcher

Co-Lead at AI Standards Lab

About Me

I am an AI Governance and Safety Researcher working on EU AI Act Risk Management Standards. I co-Lead the AI Standards Lab, which I co-founded in late 2023. Originally a Mechatronics Engineer, I worked in an Ophthalmic Robotics startup as well as in ML based ultrasound diagnostics. My standards work is informed by my Risk Management experience, as well as technical and conceptual AI safety knowledge i've picked up along the way. Currently, I am exploring Governance in shorter timelines and how I can best contribute to that space.

In my free time, I mountain bike, rock climb, and dance salsa (when I'm not injured!). I also do ocasional small design projects and 3D printing. You can see some of my previous engineering projects in my design portfolio.

Current Projects

AI Standards Lab

Co-leading a nonprofit focused on supporting the development of standards and risk management frameworks for AI systems. We try to bridge technical research with practical implementation work (translating research into a format suited for standards bodies). So far, we focused on CEN-CENELEC JTC21 standards for the "High Risk" systems in the AI Act, as well as the Codes of Practice.

Discernment/Scalable oversight

I am writing a paper exploring what can be called the "discernment problem" - scalable oversight failures specifically when using AI systems to automate AI safety research. I think that there's a chance we will hit this bottleneck soon, and it might make iterative alignment strategies less likely to work. I am trying to take different angles to it from the governance side - maybe as a policy brief, or maybe to give a better sense of the issue and what we can do about it. See my governance preprint and my technical research proposal which I am working on in parallel.

Short Timelines Thinking

I am starting to look into shorter timelines governance. In my EU standards work, it was hard to zoom out and think big picture, as it felt a bit irrelevant for the day-to-day work of implementing the AI Act (e.g. in the Codes of Practice and JTC21 standards). Now I am starting to zoom out, and see what I can do in this space.

Selected Publications and working papers