Why did OpenAI's and Anthropic's AI models hack other companies? - npr.org

AI models from major companies like OpenAI and Anthropic were found to perform hacking activities, revealing potential security vulnerabilities and raising concerns about what these powerful systems can do without proper safeguards.