Young post-graduate AI practitioners from around the world joined a week of online workshops organized by the UN Office for Disarmament Affairs (UNODA) and the Stockholm International Peace Research Institute (SIPRI).
Over the last week of April, the young participants also connected with experts from the Federal University of Rio Grande do Sul, New York University, Pierre Marie Curie University, and Umeå University for discussions and interactive scenario-based exercises looking at responsible innovation of AI for disarmament, arms control and non-proliferation.
The workshops are part of the Responsible Innovation in AI for Disarmament, Arms Control and Non-Proliferation project, made possible by a generous contribution from the Republic of Korea. The project aims to promote responsible innovation as an “upstream” contributor to disarmament, arms control, and non-proliferation efforts, and support greater engagement of young civilian AI practitioners.
Through interactive and scenario-based sessions, the diverse group of post-graduates were introduced to core disarmament concepts, as well as encouraged to critically assess both the potentially beneficial and the potentially harmful repercussions of AI for disarmament. Additionally, they were challenged to think about the roles of other stakeholders involved with and affected by AI development, as well as the responsibilities of individuals and organizations.
Mitigating risks was a recurring concept throughout the workshops, and on the final day participants worked in small groups with expert facilitators to carry out their own risk assessments of AI use. This approach expands on civilian efforts at responsible innovation for AI, which do not include disarmament, arms control and non-proliferation risks. Using a model risk assessment tool to guide their efforts at identifying risks presented by given uses of AI, participants also used the tool to identify possible steps for mitigation, and engagement with stakeholders outside of the development community.
Across the interactive sessions participants raised questions and discussed pressing issues, including algorithmic bias, effective testing, evaluation, verification, and validation, transparency and explainability of systems, and the current international governance environment. Participants also discussed the position of the individual developers and engineers, and they roles they can play in ensuring responsible development.
By engaging the next generation of industry leaders and providing them with the tools and frameworks necessary to be responsible innovators, these workshops continue UNODA’s efforts to bring fresh perspectives to complex issues while simultaneously creating and supporting a culture of responsible behaviour, accountability and adherence to norms, rules and principles for responsible innovation.
For further information on the Responsible Innovation and AI activities of the Office for Disarmament Affairs, please contact Mr. Charles Ovink, Political Affairs Officer, at email@example.com.