Heading backwards to go forwards in Military AI
This panel session will provide a series of diverse perspectives on military AI, starting with myth busting and progressing to responsible deployment and use. First, AI has been in use within the military for decades so we must ask ourselves what has really changed and what does this mean? Then we must start at the right hand side with our military customers to understand what exactly they need to make legal and ethical military decisions, weighing up the military necessity versus the risks of using AI enabled systems. From here we must work backwards to drive system design, decision making tools, training, and the implementation of ethical principles. Ultimately, military users of AI need to be able to quantify mission assurance and the risks for each unique operational context (i.e., mission, environment, constraints) to make the best possible, flexible, and responsible operational deployment decisions. Potential solutions and challenges are discussed.