Elly Millican

Hello! I'm a researcher and long-time technologist interested in algorithmic fairness and AI accountability. I work on policy research, technical AI evaluation, and implementing ethics in real-world systems. I also help organizations navigate emerging AI regulations and standards.

Recently, I've been developing methods to evaluate fairness when LLMs summarize case documents in high-stakes contexts, including insurance underwriting and immigration decision-making.

Much of my research has been on the use of technology in border systems. I investigate privatized visa processing, facial age estimation and other biometrics, automated decision-making about travelers, and the grassroots technical responses to these measures. I combine ethnographic and technical methods in my work.

the wolfson meads at twilight, with a rising moon

work

  • Fairness Testing for LLM Document Processing
    I was honored to be invited to give a lecture on counterfactual auditing of LLM document systems at the Oxford Internet Institute in March 2026. These are the slides from that lecture. /li>
  • The Visa Empire: Borders as a Business
    I was a contributing investigator on this year-long Lighthouse Reports investigation into VFS Global, published in May 2026 with 14 media partners.
  • Asylum by Algorithm
    I contributed to this Lighthouse Reports audit and inquiry into the UK Home Office's use of facial age estimation on asylum seekers, published in June 2026 with WIRED and The Independent.