Nature

AI models that lie, cheat and plot murder: how dangerous are LLMs really?

2025-10-08

Full text

DOI https://doi.org/10.1038/d41586-025-03222-1