By: James Dargan
Anthropic is confronting two separate controversies involving the safety and security of its Claude AI systems, ranging from account theft affecting paying customers to stark internal warnings about existential risks from advanced AI.