Recent work on AI (2024-2026) by Emilio Ferrara — safety, auditing, agents, generative-AI risk, evaluation; with BibTeX