Expert Workshop on Responsible Use of AI in Military Systems
Weblogs
In the run-up to REAIM 2023, an expert workshop on the responsible use of Artificial Intelligence (AI) in military systems took place on 31 October and 1 November. One of the goals of the workshop was to gather input for the REAIM Summit to be held in February 2023.
Fifty leading experts from various countries and areas of expertise discussed the topic of the workshop, which was held at the Marine Etablissement Amsterdam (historic naval storehouse). Speakers included Lieutenant Colonel Martijn Hädicke of the Royal Netherlands Army, who gave a plenary lecture on the Robots and Autonomous Systems (RAS) unit’s innovation programme. Diane Staheli (Chief of Responsible AI, Chief Digital and AI Office) spoke about the US Department of Defense’s activities in the field of responsible AI.
During breakout sessions, the participants discussed dilemmas regarding the responsible development and use of AI in military systems. Research agendas for various disciplines (legal, ethical, AI, human factors, industry and policy) were also formulated.
Key outcomes of the workshop
1. Trust is important. If AI is not understood, the system will neither be trusted nor used. On the other hand, misunderstanding can also lead to overconfidence and irresponsible use of AI in the military domain. New methods are needed to measure trust. It is also necessary to develop training courses to familiarise military personnel with AI.
2. The use of AI goes beyond weapon systems. Other application areas are involved too, including logistics and maintenance, decision support, early warning systems (e.g. in cyber/AI security), business operations and security.
3. There should be more focus on AI development upfront, rather than regulation after the fact. A transdisciplinary approach is crucial: from design to maintenance and from training to doctrine development and ethics.
Input for the REAIM Summit
The successful two-day workshop provided valuable input for the REAIM Summit. It is also a good starting point for further discussions and research in the area of responsible use of AI in military systems.