The use of Artificial Intelligence (AI) in the development of safety-critical systems poses fundamental challenges for industry and research alike. On the one hand, the application of AI-based methods in virtually all phases of systems engineering promises enormous productivity gains, leading to faster system development, shorter time-to-market, and cost reductions through the elimination of (expensive) human labour. On the other hand, particularly for safety-related system components, questions of quality assurance, suitable testing procedures, integration into the safety case, and – not least – whether AI can generate new, creative and innovative approaches in systems engineering at all without human assistance, remain largely unanswered.