Contact · Researchers
For researchers
Working on AI safety, human-computer interaction, accessibility or evaluation methodology? We'd like to hear from you.
Tell us what you're studying
Send your research question, the people or behaviours you want to study, and what you would need from us. Include any timing, consent or data requirements so we can discuss what is practical.
Understanding AI's Confidence
We're interested in how to best identify when AI is answering questions beyond its knowledge or remit. Whether it's using classification models, instructions-based guardrails or other techniques, we want systems to answer well when they can, but not overstep when they can't. This is an area of active research, so if you're interested in that area we'd love to hear from you.