Cao, Y. (2026) “Failures to Surface Harmful Contents in Video Large Language Models”, Proceedings of the AAAI Conference on Artificial Intelligence, 40(42), pp. 35331–35339. doi: 10.1609/aaai.v40i42.40841.