Compared to other global powers, the European Union (EU) is rarely considered a leading player in the development of artificial intelligence (AI). Why is this, and does this in fact accurately reflect the EU’s activities related to AI? What would it take for the EU to take a more leading role in AI, and to be internationally recognised as such?
Today, the prevailing practice in machine learning is to train a system on a training data set, and then test it on another set. While this reveals the average-case performance of models, it is also crucial to ensure robustness, or acceptably high performance even in the worst case. In this article, we describe three approaches for rigorously identifying and eliminating bugs in learned predictive models: adversarial testing, robust learning, and formal verification.