Leverhulme Centre for the Future of Intelligence

Last month I joined a diverse mix of user experience researchers and practitioners in tackling a seemingly insurmountable question: What does it mean to design AI interfaces responsibly? We gathered as part of the CHI, Human Factors In Computing, conference to explore this question from a variety of vantage points. Participants across academia, industry, and policy, brought theoretical frameworks and critiques as well as the constraints, tradeoffs, and implementation challenges that shape this work in practice. Some of the topics we explored included: • Supporting user agency • Treating transparency as a design responsibility • Designing for “exponential pathways” • Understanding friction as a feature, not a bug • Reckoning with the realities of scale • Keeping focus on positive change • Resisting the urge to humanize AI You can read more about the session and these topics here: https://lnkd.in/gUbFFMgk

I am thrilled to share that next week I will graduate from Harvard College, Phi Beta Kappa, with double concentrations in Government and Applied Mathematics. I could not be more grateful for my time at Harvard—I have learned so much, formed friendships I will cherish for a lifetime, and had the privilege of learning from extraordinary mentors, professors, and peers. This year, I wrote a senior honors thesis in the Department of Government titled “Stare Indecisis Revisited: Attitudinal Models in the Roberts Court and an Examination of the Supreme Court’s Precedent on Precedent.” My thesis examined one of the central puzzles in judicial decision-making: why do courts overturn their own precedent? For this work, I was awarded the Philo Sherman Bennett Prize for the “best essay discussing the principles of free government… that is, in the opinion of the judges, worthy of publication.” I also received highest departmental honors and the Charles Joseph Bonaparte Prize. Next year, I will attend the University of Cambridge as a member of Darwin College, where I will pursue an MPhil in the Ethics of AI, Data, and Algorithms at the Leverhulme Centre for the Future of Intelligence. My research will focus on the intersection of artificial intelligence and judicial decision-making. To support my studies, I was selected as part of the inaugural cohort of the Spärck Scholarship, awarded to “young people with exceptional talent in AI.” After my time in England, I will attend Harvard Law School as a member of the Class of 2030. Recently, I was also honored to receive the Endicott Peabody Saltonstall Prize, awarded to the Harvard senior entering Harvard Law School who is considered “best fitted, by intellect, character, and physique, to be influenced by Saltonstall’s example and in turn to influence others.” Thank you to everyone who made these last four years so meaningful. I could not be more excited for four more years of living in Cambridge—whether England or Massachusetts.

First Ethics Report: Building Trustworthy AI for Labour Market Policies 🌍 Ethics cannot be an afterthought in AI-supported labour market policy.  The EU-ALMPO project has published its First Ethics Report, the first formal ethics report of the project. Authored by Tomasz Hollanek, PhD, University of Cambridge, acting as Ethics Advisor for EU-ALMPO, the report reviews how ethics, data protection, accountability, stakeholder engagement, and human-centred AI are embedded across the project’s work. It highlights the project’s ethical foundation, built around:  ▪ fundamental rights and EU values  ▪ transparency and accountability  ▪ participatory co-design  ▪ context-specific implementation across partner countries  ▪ safeguards for data protection and AI-supported tools The report also identifies key areas for continued attention, including ethics training, stronger reporting channels for ethical concerns, and the careful assessment of AI-related risks. For EU-ALMPO, responsible innovation means ensuring that AI-supported tools for Active Labour Market Policies are not only effective, but also fair, transparent, and trustworthy. 👉 Read the report on our website: https://lnkd.in/duFjDBnY #EUALMPO #HorizonEurope #ActiveLabourMarketPolicies #TrustworthyAI #EthicalAI #AIForPolicy #LabourMarketInnovation #DataGovernance 

Excited to share that I officially graduated from my MSc in the Social Science of the Internet at the Oxford Internet Institute, University of Oxford this past weekend (although I completed it last August!) I was incredibly grateful to receive a distinction on my thesis, which explored how we think about and govern AI under the EU AI Act, and I had the opportunity to present some of this work at PAIRS - Participatory AI Research & Practice Symposium in the India AI Impact Summit earlier this year in February. Being back in Oxford this weekend reminded me how much I learned in just one year. I tried rowing for the first time, sharpened my R and Python skills and wrote about agency in the AI/social media era and the implications of platform bans. It was a year that pushed me intellectually, and gave me friendships and memories that will last a lifetime. A huge thank you to my two amazing supervisors, Fabian Stephany and Johann Laux, who challenged me to approach my thesis from interdisciplinary perspectives. Together, they created a space where I could bridge law, policy, and computational social science in ways that have fundamentally shaped how I think about AI governance today. I'm incredibly grateful for their guidance, support, and patience throughout the year. Since completing the MSc, I’ve been lucky to work with some amazing organisations and people, including Dr Eleanor Drage, Tomasz Hollanek, PhD, and more recently Beryl Pong at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge. I’ve been researching how we might reimagine AI regulation and AI ethics, specifically looking at high-risk AI applications under the EU AI Act, cross-cultural approaches to AI ethics, and the lived psychological dimensions of drone warfare. For the past couple of months, I’ve also been working as an AI Research Fellow with Reframe Venture, a non-profit that works with over 500 VCs and 110 LPs to advance responsible investment practices. My work focuses on sector-specific AI risks in education and healthcare, helping investors build expertise and better evaluate and respond to emerging AI challenges. Alongside this, I’ve been conducting research with the amazing team at Digital Trust Council, where we’ve been exploring what a certification framework for trustworthy AI could look like in practice. I’ve been spending a lot of time thinking about how we understand AI risk from political, social, and institutional perspectives, and I feel incredibly lucky to be able to work across academia and industry with such thoughtful and passionate people. Lots of exciting research and papers coming soon! If you want to chat about AI regulation/ethics/safety, my DMs are open!

From 25-26 June 2026, we will be hosting the "Critical Drone Studies: Drones in Society, Politics, and Culture" conference at the University of Cambridge. I wanted to reflect on why we're putting together this event with my colleagues. There are plenty of critiques levelled against academia around how slow we are to think and publish, and how, to understand "new" emerging technologies, readers ought to go elsewhere for the latest developments. That is true, but that is not necessarily what motivates or preoccupies all researchers. Some are not chasing the newest developments, and some are not trying to predict the future (because the future still open and can be shaped). Some are not interested in feeding the arms race rhetoric of "innovating first" or else being left behind (behind whom? behind what? Who decides this? What are we arming for?). This is an event about drones that hopes to foreground some abstract questions, yes, but ones that are still very tangible. Where do we want drones in politics, society, and culture, and why? What do current takes on drones obscure about longer histories of aerial violence and security? How do we balance talking about drones in terms of governance, technicities, and regulation alongside issues of creativity, prosthesis, motivations, and access -- in terms of where we think "unmanning" is appropriate or desirable, and for whom, and if at all? (because there are always humans, and non-human beings, involved) If these kinds of questions sound up your alley, come join us. Spots to attend are still available and sponsored by my #UKRIFLF from UK Research and Innovation. We have an incredible group of researchers from around the world presenting, and we're also showcasing some artworks engaging critically with the so-called drone age. Details here: https://lnkd.in/e8NykjvA Leverhulme Centre for the Future of Intelligence Amy Gaeta Márjory Da Costa-Abreu Richard A Carter Joanna Tidy