Koyanagi is a researcher at the University of Cambridge, working on the governance of space, nuclear, and cyber systems — how authority is constructed, how norms break down, and what it takes to rebuild legitimacy. She is Team Leader of SGAC's 2026 Space Policy Project Team (Japan), a delegate to the JAXA International Space Education Board (2026), and a mentee in the VCDNP 2026–2027 Young Women in Non-Proliferation and Disarmament Programme — one of twenty selected worldwide.

As a Cambridge Future Scholar, she conducts supervised research under Prof. Harald Wydra on competing lunar governance frameworks. Her broader work spans dual-use space infrastructure, cross-domain escalation risk, and gaps in responsible-behaviour norms across AI, space, and cyber governance — grounded in six years of field research in post-disaster Fukushima.

Her work is forthcoming at the International Astronautical Congress 2026 (Antalya, Türkiye), with three accepted papers on dual-use neutralisation, conditional openness in lunar governance, and mythic naming in competing lunar orders.

Alongside her research, she builds two open research platforms — Faultline (strategic infrastructure risk) and Govern the Human (how AI reshapes the human subject). Both advanced to the prototyping phase of the Global Challenge to Build Trust in the Age of Generative AI — a G7-originated initiative led by the IEEE Standards Association, the OECD, and AI Commons, with UNESCO as a founding partner — among only six East Asian teams of eighty selected worldwide.