Adrian Arnaiz-Rodriguez

Adrian Arnaiz-Rodriguez

Research Scientist

Apart Research

I am a Research Scientist at Apart Research, studying manipulation risks in frontier AI systems and developing third-party AI evaluations. I’m also a member of the ELLIS network.

I previously worked as a postdoctoral researcher at ELLIS Alicante, where I also completed my PhD (cum laude) under the supervision of Nuria Oliver. During my PhD, I was a visiting researcher at MPI-SWS, hosted by Manuel Gomez-Rodriguez as part of the ELLIS PhD & PostDoc Program.

⚖️+🤖+🏛️ My research focuses on building trustworthy AI systems for high-stakes decisions and social networks, and on exploring how these technologies intersect with AI regulation and policy-making, understanding AI as a sociotechnical system.
🕸️ I am also particularly interested in graph neural networks: understanding their theoretical and practical limitations, improving performance and interpretability, and analyzing fairness and robustness in networked settings. I co-organized the ICML 2024 tutorial on Graph Learning with Ameya Velingker, and served as General Chair of LoG 2025 and Associate Chair of ICML 2024.

👆 Learn more about my journey

🎓 During my PhD and postdoctoral research, I focused on technical and interdisciplinary aspects of Trustworthy AI: algorithmic fairness, human-AI collaboration, AI safety in LLMs and multi-agent systems, and their alignment with AI regulatory frameworks and labor law. I also studied the limitations of graph neural networks.

Thesis   Publications   Tutorials

🌍 During my doctoral research, I engaged in project-based collaborations with industry and institutional partners, including Intel Labs (RESUMAIS project) and Bank Sabadell, addressing responsible AI integration in operational settings. At the European level, I contributed to the research and innovation project ELIAS. At the national level, I participated in the development of Spain’s Charter of Digital Rights, contributing to policy discussions on AI governance, fundamental rights, and algorithmic accountability.

📢 Beyond research and policy work, I played an active role in dissemination, outreach, and organization efforts within the research community.

🔙 Before my PhD, I worked on network science and ML for life sciences at UOC with Baris Kanber (UCL) in my MSc thesis, and at the Admirable lab at the University of Burgos with a research scholarship in my BSc thesis.

Interests
  • AI manipulation and AI evaluations
  • Trustworthy AI for high-stakes settings and social networks
  • GNNs and network science
  • AI regulation and policy-making
Education

  • PhD in Computer Science, 2025

    ELLIS Alicante (exchange at MPI-SWS)

  • MSc in Data Science and AI, 2021

    Open University of Catalonia

  • BSc in Computer Science, 2019

    Burgos University

Tutorials

Contact