Anthropic changes Safety guidelines, will now train AI model even if safety not guaranteed

Facing competition, Anthropic has revised its safety framework. Earlier it had decided that it would not train an AI model if ...