This is a living document in which senior AI researchers leading an ICAI Lab in Amsterdam express what they see as the core values underlying their research. We welcome feedback and will update (and track changes) as we go forward.
Like all science, AI is not practiced in a vacuum, but in interaction with society. We therefore want to be explicit about our core values and the guidelines we follow to uphold these values. They apply to our researchers, the contents of our research as well as to the choice of our partners. We continuously improve our core values and guidelines to ensure they reflect the expectations of today’s society.
- We strive for a positive social impact and economic value of our research (“AI for social good“), as expressed in the Sustainable Development Goals of the United Nations, and as endorsed by the Dutch universities. We subscribe to the Ethics Guidelines for Trustworthy AI as formulated by the High-Level Expert Group on AI of the European Commission.
- We are aware of the transformative competencies of AI in shaping society, creating new values and reconciling tensions and dilemmas. We contribute to social awareness about how the world is changing and will change under the influence of AI, we initiate and conduct the social debate about the role of AI in society.
- We defend academic freedom and independence as formulated in the Dutch code of conduct, and we publish our scientific results under open access, with the exception of trade secrets and IP protection, on which we make transparent agreements.
- We are transparent about where we stand academically, where we are going and how. We are open about dilemmas we may face, how we will deal with them when they arise and what choices we make.
- We regard the training of scientific talent as one of our core objectives.
- We engage in dialogue with and build bridges to both emerging and established scientists in other countries and cultures, contributing to national and international communities for knowledge sharing, collaboration and innovation.
- We strengthen our local ecosystem for AI research and its applications, in collaboration with universities, industry, government and social institutions.
To uphold our core values in practice, we rely on guidelines for how to act, whether as an individual researcher or as a lab. Our current guidelines include the following:
- For research involving human subjects or privacy-sensitive data we consult the Ethical Committee for Information Sciences before the start of our research. Where relevant (for example on issues of social impact, national security or international relations) we consult other organizations, such as municipalities, regulatory authorities, advisory councils, NGOs and other societal stakeholders.
- In relevant cases we inform other ICAI labs, the ICAI advisory board and the ICAI member meeting about social and ethical considerations.
- During the research we regularly evaluate how the current research relates to the above values and goals, and inform the other ICAI labs, the ICAI advisory board and the ICAI member meeting if necessary.
Marieke van Erp (Cultural AI Lab)
Theo Gevers (ATLAS Lab)
Sennay Ghebreab (Civic AI Lab)
Paul Groth (AIRLab Amsterdam, Discovery Lab)
Laura Hollink (Cultural AI Lab)
Frank van Harmelen (Discovery Lab)
Jacco van Ossenbruggen (Civic AI Lab, Cultural AI Lab)
Maarten de Rijke (AIRLab Amsterdam)
Cees Snoek (AIM Lab, ATLAS Lab, QUVA)
Max Welling (Delta, QUVA)
Marcel Worring (AIM Lab)
- Original version: December 14, 2020