Artificial Intelligence (AI) presents transformative potential for science academies, offering tools to enhance the various tasks and processes carried out by these organisations. They range from research synthesis to administrative functions, scientific advisory processes to marketing and engagement with academia. This event explores how AI can support the mission and activities of science academies while fostering inter-academy cooperation.
As academies navigate the integration of AI, critical ethical challenges arise – including algorithmic bias, transparency and accountability – which must be addressed to maintain the integrity and credibility of scientific institutions. The event will examine how AI can be responsibly leveraged for different functions, with a particular focus on scientific processes and the nuances of science advice. Key considerations include the need for targeted training to equip staff and stakeholders with the skills to evaluate and use AI tools effectively and ethically.
This event will also reflect on the suitability of current AI solutions for supporting evidence-informed policy processes and the importance of safeguarding core values such as independence, inclusivity, and rigor. Through cross-academy dialogue, this is an opportunity to chart a responsible path forward for AI in science academies.
The discussion will also reflect the work carried out by the SAPEA consortium over the past year in establishing guidelines for the use of AI tools within the Scientific Advice Mechanism to the European Commission.
Programme
Please note that All times are in Central European Time (CET). If you are attending in-person in Estonia, the time is one hour later.
09:30–10:00
Registration, welcome and coffee
10:00-10:45
Keynote Address Jaak Vilo, Vice-President of the Estonian Academy of Sciences
The role of science academies in the 21st century: purpose and cross-border cooperation
- Tasks and activities of science academies
- Challenges and opportunities in changing world
- Interaction with research and the academic world
- Can the use of AI tools be an effective tool to support mission of science academies?
Followed by Q&A moderated by Hanna-Leena Pesonen, EASAC Vice-President
10:45-11:00
Coffee break
11:00–12:00
Panel Discussion | AI in Science: challenges and opportunities
Increasing accessibility of Generative AI and other machine learning tools has led scientists across different disciplines to incorporate them into their researchers. Beyond facilitating ground-breaking discoveries, AI is also transforming the daily academic work of scientists.
AI has the potential to accelerate discovery and innovation, automate workflows, enhance output dissemination. However, researchers must also address bias, respect principles of research ethics and integrity, and deal responsibly with issues surrounding reproducibility, transparency and interpretability.
How to balance challenges and opportunities facing academics when it comes to using AI tools? How to safeguard public trust in research, and ensure personal responsibility? How to embed responsible AI in wider research integrity frameworks? What sort of intervention should legal frameworks have in this field? What role for science academies in supporting their members on responsible use of AI?
Moderation: Georg Bërveniku-Brunner, EASAC Executive Director
Participants: Helen Eenmaa (Tartu University), Maarja Kruusmaa (Tallinn University of Technology)
12:00–13:00
Lunch break
13:00–13:15
Presentation | Guidelines and procedures for the use of AI in the SAM
Frederico Rocha, SAPEA Scientific Policy Officer (EASAC)
13:15-14:15
Panel Discussion | The use of AI tools in scientific advice for policymaking
Science for Policy (S4P) refers to the use of scientific evidence, expertise, and knowledge to improve policymaking. In this context, science helps to identify and understand policy challenges; identify and evaluate interventions and policy alternatives; make informed decisions; design effective interventions.
Considering the increasing pace of the policymaking processes and the emergence of polycrisis, discussion has flourished on whether advances in AI technology can increase the capacity of science advice stakeholders in supporting policymakers when it comes to information regarding rapidly evolving issues, emergencies, and debates.
Is science for policy ready to embrace AI amidst challenges relating to potential bias, errors and hallucinations? Can policymakers already do without human-led scientific advice? How to measure the risks, safeguard public trust and ensure accountability? Can scientific advice stakeholders, such as science academies, build a responsible path for using AI in science for policy and bring policymakers along with them?
Participation: Justyna Bandola-Gill (University of Birmingham), Alar Karis (President of Estonian Republic 2021-2026), Frederico Rocha (SAPEA)
14:15–14:30
Coffee break
14:30–15:30
Panel Discussion | What role for AI in advancing the mission of science academies?
Science academies have long been at the core of knowledge provision, and their role has evolved alongside societal change. These institutions comprise some of the best and most renowned scholars, responsible from substantive advances in science. Many of them also hold an important mission in supporting evidence-based policymaking.
As in any other sector, organisations are reflecting on how to embed the use of AI in their daily activities. AI tools are seen as potential vehicles to increase efficiency, and address resource shortages. At the same time, the breath of scientific, administrative and communication tasks within science academies require careful reflection when it comes to risk, in-house knowledge and institutional support.
Considering the challenges and opportunities set out the previous panel discussions, how can science academies make best use of AI to deliver their own activities? How can staff make informed decisions on whether to use AI in fulfilling the academies’ daily activities? How support structures should be in place for a successful uptake of AI? Should academies support their members in their own AI literacy?
Moderation: Simon Schneider, SAPEA Coordinator
Participation: Linda Lammensalo (Finnish Academy of Science and Letters), Tarmo Soomere (Estonian Academy of Sciences), Christian Weidlich (German National Academy of Sciences Leopoldina)