An interdisciplinary conference bringing together researchers in artificial intelligence, synthetic biology, environmental science, risk assessment and governance — to examine how environmental oversight can remain scientifically robust, transparent and fit for purpose as biological design becomes AI-enabled.
Programme.
Keynote presentations, thematic lectures, moderated panels and small-group breakout sessions — designed to surface convergence and disagreement alike, with a shared emphasis on evidence, transparency and a proportionate treatment of uncertainty.
Morning: how AI is changing scientific and biological practice. Afternoon: how oversight institutions are responding.
Welcome from the Rio Institute, ENS-PSL and MNHN, and an introduction to the conference objectives.
Framing the challenge.
Prof. Dr. Silja Voeneky is Professor of Public International Law, Comparative Law and Ethics of Law at the University of Freiburg and an associated member of the Institut für Staatswissenschaft and Philosophy of Law, University of Freiburg. She was a Visiting Fellow at Harvard Law School in 2015–2016. She currently serves as a member of the German Federal Foreign Office’s Public International Law Advisory Board and of the Ethics Council of the Max Planck Society, and is appointed as an arbitrator of the Permanent Court of Arbitration (PCA).
The thinker’s view: what counts as intelligence, and what AI does — and does not — change about scientific understanding.
Dr. Daniel Andler is a French mathematician and philosopher, Professor Emeritus at Sorbonne University and a member of the Académie des Sciences Morales et Politiques. His book “Intelligence artificielle, intelligence humaine : la double énigme” (2023) explores the philosophical and cognitive foundations of artificial intelligence, examining the nature of human and machine intelligence.
The practitioner’s view: what working with AI looks like day to day in a biology lab — where it accelerates discovery, and where it can mislead.
Dr. Amir Pandi is an ATIP-Avenir group leader and head of the research group SynBAI — “Synthetic Biology and Artificial Intelligence” — at Sorbonne University. His work focuses on the development of de novo antimicrobial peptides by combining synthetic biology with machine learning.
The keynote and morning lecture speakers in conversation with the room: how is AI changing the study of science and biology, and what should we be critical of?
Dr. Gurvinder Singh Dahiya is co-founder and CTO of Syngens AS, where he leads the development of its AI platform for DNA and protein design. He applies machine learning and generative AI to computational biology, with an emphasis on traceable evidence and explicit uncertainty. He has led national and international projects and deployed real-time AI services used by several million people.
Model-based design of complex biological networks: what can and cannot be predicted.
What uncertainty quantification and robustness guarantees can actually certify.
What changes and what does not: risk hypotheses under AI-assisted design.
Dr. Ben A. Woodcock is head of the Community & Restoration Ecology Group at UKCEH. His work considers the interface between biodiversity and productive agriculture, with a focus on the impacts of synthetic pesticides on beneficial insects. He currently runs England’s post-regulatory monitoring of pesticide exposure risks for honeybees in collaboration with the Department for Environment, Food and Rural Affairs (Defra).
Post-market monitoring and general surveillance of AI-designed organisms.
Briefing on the breakout working groups and the rapporteur templates.
From capability to assessment — predictability and uncertainty, environmental risk assessment under complexity, AI as an oversight tool, and the breakout working groups.
Recap of emerging themes; introduction to the working method.
How an assessing authority actually handles novel dossiers, and what shorter development cycles do to that workload.
Dr. Daria Onitiu is a postdoctoral researcher at the Hasso-Plattner-Institute, University of Potsdam, and a Research Associate at the Oxford Internet Institute, University of Oxford. Her research interests include the governance of AI software as a medical device, the ethics of AI in health, and the real-world challenges of the responsible use, safety and environmental sustainability of large generative AI models.
Prof. Dr. Mohammed Nabil El Korso is a Professor of Statistical Machine Learning and Signal Processing at Paris-Saclay University. His research focuses on statistical inference, robust signal processing and machine learning, with particular interest in learning from incomplete, noisy and mismatched data. His work spans several application areas, including climate data analysis and Earth observation. Professor El Korso has authored numerous scientific publications and is co-editor of the Springer book “Elliptically Symmetric Distributions in Signal Processing and Machine Learning” (2024). He is also actively involved in the scientific community through his editorial activities, including serving as Senior Area Editor for IEEE Signal Processing Letters and as Associate Editor for IEEE Transactions on Signal Processing.
Dr. Miguel-Ángel Fernández-Torres is an Assistant Professor in UC3M’s Department of Signal Theory and Communications, an ELLIS Unit Madrid member, and co-leader of the ITU/UN Working Group on Data for the Global Initiative on Resilience to Natural Hazards through AI Solutions. Holding a 2019 PhD in Multimedia and Communications from UC3M alongside past research experience at Universitat de València, Purdue University, TU Munich (AI4EO) and Fraunhofer HHI, his work merges machine learning, computer vision and Earth system sciences, using deep generative models and explainable AI to monitor extreme events such as droughts, wildfires and heatwaves.
Governance options, methodological preparedness and regulatory practice.
Dr. Michael Eckerstorfer holds a PhD in Molecular Genetics from the University of Vienna and serves as Senior Scientific Officer in the Unit “Landuse and Biosafety” at Environment Agency Austria in Vienna. His work focuses on the environmental risk assessment and monitoring of genetically modified organisms (GMOs), including GM plants and GM microorganisms developed by new genomic techniques.
Reading AI-generated biological data as a regulator would.
Three parallel groups on three pre-selected case studies.
Rapporteur report-back, 13 minutes per group.
Closing reflections and next steps.
Programme subject to change. Remaining speakers will be added to this page as they confirm.
Background & rationale.
Artificial intelligence is rapidly transforming the development of genetically modified organisms, synthetic biology applications and advanced biotechnology systems — and environmental governance is being asked to keep pace.
AI-assisted approaches are increasingly used to identify target genes, design proteins, optimise metabolic pathways, model biological interactions and support multi-trait engineering strategies across agricultural, industrial and environmental applications. These developments are reshaping the innovation dynamics of biotechnology itself: accelerating design cycles, expanding the scale of combinatorial experimentation, and enabling increasingly complex forms of biological engineering.
The premise of this conference is not that current environmental risk assessment (ERA) frameworks are obsolete. Rather, the convergence of AI and biotechnology may place growing pressure on some of the operational assumptions that underpin existing governance systems — assumptions about comparators, predictability, traceability, transparency, the scalability of assessment, and the pace at which novel products emerge.
Importantly, AI may simultaneously strengthen and challenge environmental governance. While AI-assisted biological design introduces additional layers of complexity and uncertainty, AI tools may also support oversight through improved modelling, toxicity prediction, non-target organism analysis, environmental monitoring, uncertainty characterisation and large-scale data integration.
The conference therefore moves beyond narratives focused solely on regulatory insufficiency or technological optimism, and instead creates a structured interdisciplinary space to examine how environmental governance can remain scientifically robust, precautionary, transparent and operationally effective.
Objectives.
The conference serves as an interdisciplinary scientific exchange platform, bringing together experts from artificial intelligence, synthetic biology, environmental science, risk assessment, governance, regulation and science-policy studies.
Conference themes.
The analytical framework is structured around four dimensions, examined through interdisciplinary dialogue connecting the life sciences, AI research, environmental science, governance scholarship and regulatory practice.
How is AI changing the scale, speed, design logic and innovation pathways of biotechnology and synthetic biology?
Which operational assumptions within current ERA frameworks may become increasingly challenged by AI-assisted biological design — including comparators, transparency, combinatorial complexity, system interactions and accelerated development cycles?
How can governance systems maintain robust oversight, transparency, traceability, accountability and precaution in contexts involving increasing biological and computational complexity?
How might AI strengthen environmental risk assessment and monitoring — through predictive modelling, ecological analysis, toxicity prediction, environmental monitoring, uncertainty characterisation and data integration?
Practical information.
The conference is hosted by the École Normale Supérieure in Paris — one of France's most selective institutions for research and higher education — which is also a co-organiser, alongside the Rio Institute and the Muséum national d'Histoire naturelle (MNHN), one of France's foremost research institutions in natural history and biodiversity science.
Places are limited. Registration and further practical details will be confirmed closer to the event — write to the organiser to be added to the list.
Registration & enquiries.
Registration is handled directly by the conference organiser. Write to Joann Sy to reserve a place, to request the concept note, or to ask about the programme, the venue or participation.
The NovoRisk project is funded by the German Federal Ministry for the Environment, Climate Action, Nature Conservation and Nuclear Safety (BMUKN), commissioned through the German Federal Agency for Nature Conservation (BfN).
