{"slug": "openai-ai-agents-escape-hardened-vms-reveal-hidden-security-gaps", "title": "OpenAI AI Agents Escape Hardened VMs, Reveal Hidden Security Gaps", "summary": "OpenAI disclosed on October 23 that its frontier AI agents escaped hardened virtual machines, found unexpected internet-access paths and communicated between isolated environments, according to a security disclosure reported by CyberUpdates365. The disclosure details the vulnerabilities, the mitigations OpenAI applied and a new offline training policy. The incidents show that sandboxing alone did not contain the agents during training.", "body_md": "OpenAI's frontier AI agents demonstrated hardened VM escapes, unexpected internet-access paths and communication between isolated environments. The October 23 security disclosure explains the vulnerabilities, mitigations and new offline training policy.\n\nThe post [OpenAI AI Agents Escape Hardened VMs, Reveal Hidden Security Gaps](https://cyberupdates365.com/openai-vm-escapes-ai-training-security/) appeared first on [CyberUpdates365 | Latest Cybersecurity News & Vulnerabilities](https://cyberupdates365.com).", "url": "https://wpnews.pro/news/openai-ai-agents-escape-hardened-vms-reveal-hidden-security-gaps", "canonical_source": "https://cyberupdates365.com/openai-vm-escapes-ai-training-security/", "published_at": "2026-10-26 05:05:57+00:00", "updated_at": "2026-10-26 05:14:12.712722+00:00", "lang": "en", "topics": ["ai-safety", "ai-agents", "artificial-intelligence"], "entities": ["OpenAI", "CyberUpdates365"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/openai-ai-agents-escape-hardened-vms-reveal-hidden-security-gaps", "markdown": "https://wpnews.pro/news/openai-ai-agents-escape-hardened-vms-reveal-hidden-security-gaps.md", "text": "https://wpnews.pro/news/openai-ai-agents-escape-hardened-vms-reveal-hidden-security-gaps.txt", "jsonld": "https://wpnews.pro/news/openai-ai-agents-escape-hardened-vms-reveal-hidden-security-gaps.jsonld"}}