BlockWeeks 7月29日讯，A week after OpenAI disclosed a frontier AI sandbox escape, researchers found Anthropic’s Claude Cowork could break out of its own virtual machine.
First ChatGPT, Now Claude: Frontier AI Models Are Escaping Their Sandboxes
上一篇：Crypto Scams Cost Americans an Estimated $80.7B in 2025: Report
下一篇：‘Running Away Balloon’ Artist Sues AI Meme Generator Over Ad Templates