Anthropic AI development report draws scrutiny from safety researchers

Third-party safety researchers released a report about Anthropic's AI development, including findings on productivity and recursive self-improvement.