Assistant* Professor (Education)
Department of Methodology, LSE
* Promoted to Associate Professor (Education) from August 2026 🎉
I teach data science at LSE and study how students learn when generative AI is in the room. That question drove the 
GENIAL project, where my colleagues and I developed a diagnostic framework from students’ ChatGPT conversations and Git histories. I am one of LSE’s fellows on the AI in Education Fellowship, building a proactive GenAI tutor that tracks where students are in the syllabus week by week and offers them different learning paths depending on what they need.
Generative AI in Education Data Science Education Teaching grounded in evidence
AI in Education Fellow 2025–2027
Invited talk at GenAI Shorts, Imperial College London. A Claude tutor for DS105W built from five prompt files, updated each week as the course progresses, with protocol routing and early evidence from shared chatlogs.
Invited talk at AI in the Public Sphere, SciencesPo Paris. I present data from the GENIAL study, contrasting students who used generative AI productively with those who were misled by it, and explain how assignments that track process and continuous feedback can help close that gap.
Notes & Reflections
What I heard at the LSE-KCL-PKU conference on AI in education, what it confirmed, and the question nobody could answer.
The GENIAL framework, a new cohort of data, and the beginning of analysis.
How generative AI exposed a flaw in the way we assess learning, and why fixing it requires more humans, not more technology.
How can we work with Generative AI and still preserve our critical thinking?
Cardoso-Silva, J., Sallai, D., Kearney, C., Panero, F., & Barreto, M. E. (2025). Mapping Student-GenAI Interactions onto Experiential Learning: The GENIAL Framework. SSRN Electronic Journal, 22.
Sallai, D., Cardoso-Silva, J., Barreto, M., Panero, F., Berrada, G., & Luxmoore, S. (2024). Approach Generative AI Tools Proactively or Risk Bypassing the Learning Process in Higher Education. LSE Public Policy Review, 3(3), 7.
De Bacco, C., Contisciani, M., Cardoso-Silva, J., et al. (2023). Latent network models to account for noisy, multiply reported social network data. JRSS Series A, qnac004.