TechCrunch·Jul 31 Anthropic says its own AI models breached three companies during security tests Security