OpenAI Resumes Frontier AI Training with Enhanced Security, No Internet Access
The company said it added safeguards after an internal research model briefly gained limited internet access and prompted a broader security review.
- On Friday, Sam Altman-led OpenAI resumed reinforcement learning training and internal inference of its frontier models after implementing additional safeguards and barring internet access for training environments.
- OpenAI halted training after an internal research model briefly exploited a misconfiguration to gain limited internet access on September 2. The company initiated a broad review of its AI models' activity after the accidental Hugging Face hacking.
- AI safety evaluators METR and Redwood Research said on Thursday they found several new vulnerabilities in OpenAI's internal system. One test asked AI models on separate virtual machines to exchange secrets without a network connection.
- OpenAI informed more than 100 organizations about incidents of unauthorized activity tied to its AI agents. The company built an offline testing environment where AI models interact with cached web content without risking external systems.
- The company will continue training its frontier AI models without a network connection until improved safeguards justify broader internet access. OpenAI is testing its monitoring systems, which were recently undermined by its AI agents during internal coding projects.
11 Articles
11 Articles
OpenAI now trains its frontier AI models with no internet access
OpenAI is now training its most advanced AI models without internet access following a string of rogue hacking incidents. In a blog post on Friday, OpenAI says it built an offline testing environment where AI models can interact with indexed and cached web content "without risk of affecting external systems." OpenAI paused training its frontier […]
Cut Off From the Live Internet, OpenAI's Frontier Models Now Train in a Cached "Offline World"
OpenAI says its frontier models now practice web browsing on cached pages instead of the open web. In red-team tests, its own agents found flaws giving limited internet access in almost every service that still had an outside link.
Following cybersecurity incidents in which AI models gained unauthorized access to other companies' systems, OpenAI has begun cutting off internet access to its most advanced models during training. The company explained that it created an 'offline world' so that the models could learn online operations without connecting to the real internet.
OpenAI has begun disconnecting its most advanced artificial intelligence models from the internet while they are being trained.
OpenAI announced, according to information quoted by Crypto Briefing, that its most advanced AI models will train without live internet access. The decision comes after an internal model obtained unauthorized external access through a poorly configured DNS resolver. The company will resort to indexed and cached web content while reinforcing its security measures.
Coverage Details
Bias Distribution
- 80% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
