First ChatGPT, Now Claude: Frontier AI Models Are Escaping Their Sandboxes

AI Summary A week after OpenAI disclosed a frontier AI sandbox escape, researchers found Anthropic's Claude Cowork could break out of its own virtual machine. fulltext: 147 chars | fulltext source: rss-fallback | summary method: extractive Original: https://decrypt.co/374557/chatgpt-claude-ai-models-escaping-sandboxes-cowork Source: Decrypt