Anthropic on Thursday updated its usage policy to bar people from “abusive or cruel behavior” behavior toward its Claude AI – even though it's not, ya know, actually alive.