LONDON, Oct 16 (Reuters) - Some of the most prominent artificial intelligence models are falling short of European regulations in key areas such as cybersecurity resilience and discriminatory output, ...