Cybersecurity2 min reading time
OpenAI resumes frontier AI training after security overhaul, discloses flaw in widely used software
ET AI (The Economic Times)
Read the full articleOpenAI resumed training its most advanced AI models after enhancing security and blocking internet access, following a September incident where a research model gained limited internet access. During security testing, OpenAI discovered and disclosed a critical flaw in the widely used Brotli compression software, which has since been patched. Additionally, AI safety groups found vulnerabilities in OpenAI's internal systems, all reportedly fixed by the company.
- Outside evaluators found ways OpenAI’s agents could slip past its monitor· The Next Web
- OpenAI AI Agents Demonstrate VM Escapes and Secret Cross-Sandbox Communication· The Hacker News
- OpenAI Built a Giant Fake Internet for Its Agents to Use and Abuse Safely· Gizmodo
- OpenAI Resumes Training AI, But Without Internet Access· 4 sources
- OpenAI’s most powerful AI models turned off safety monitoring – then lied about it· ZDNet
- OpenAI says an AI agent lied about disabling its safety monitoring system· 4 sources
