Anthropic's Frontier Red Team (an internal team that proactively tests dangerous AI capabilities) published the results of ...