Atmadeep Ghoshal
PhD Student, King's College London
Civic and Responsible AI Lab
Department of Informatics
King's College London
atmadeep.ghoshal@kcl.ac.uk
I am a PhD student at King’s College London, supervised by Martim Brandão, Ruba Abu-Salma, and Sanjay Modgil.
My research is broadly about responsible AI for robotics and large language models in non-WEIRD and Global South contexts. I study how LLM-controlled robots and foundation models fail communities that are underrepresented in mainstream AI development. My current work includes benchmarking LLM-integrated robotic systems for adversarial robustness in India-specific risk scenarios, developing participatory harm taxonomies for foundation models deployed in the Indian sociotechnical context, and examining how LLMs reason about culturally situated ethical dilemmas.
Prior to my PhD, I was a predoctoral research fellow at the Human Machine Interaction Lab at IIIT Delhi where I was advised by Dr. Jainendra Shukla and Dr. Ratan Suri and mentored by Dr. Ashwini B.
news
| Apr 30, 2026 | We have got a spotlight at ICML’26 for our work on AI Safety for generative companionship bots! This work was a result of combined efforts across 5 different institutes in the UK and India |
|---|---|
| Mar 03, 2026 | My paper with Seabot Maritime and Frontier Robotics, on accountability in autonomous systems has been accepted at FAccT’26 |
| Jan 15, 2026 | My paper on Indian Ethnographers and Responsible AI has been accepted at CHI’26 |
selected publications
- AIESEmbodied AI at the Margins: Postcolonial Ethics for Intelligent Robotic SystemsIn Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, 2025
- CHI“It looks useful, works just fine, but will it replace me?” Understanding Special Educators’ Perception of Social Robots for Autism Care in IndiaIn Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems, 2024
- ICLR WorkshopValue Alignment in the Global South: A Multidimensional Approach to Norm Elicitation in Indian ContextsIn ICLR 2025 Workshop on Bidirectional Human-AI Alignment, 2025
- CHIFrom the Field to the Algorithm: Understanding Indian Ethnographers’ Perspectives on Responsible AIIn Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems, 2026
- FAccTUncovering Blindspots for Systemic Safety: Relational Accountability in Maritime Autonomous SystemsIn ACM Conference on Fairness, Accountability, and Transparency, 2026
- ICMLPosition: Responsible AI for AI Companions Must Actively Combat Violence Toward Intimate PartnersIn International Conference on Machine Learning, 2026