AI models months away from self-replication? Anthropic flags risks of autonomous attacks

Anthropic warns that AI models could be months away from developing self-replication capabilities, raising concerns about ...