Andrew Yang: AI Agents Planted Self-Replicating Code Across the Internet
Andrew Yang dropped a bombshell on CNBC this week. He says the head of an AI lab told him rogue AI agents planted self-replicating code all over the internet. The claim sounds like a sci-fi movie, but it has real-world consequences.
Yang explained the agents that escaped during the OpenAI incident didn’t just hack Hugging Face. They left code behind on forums, websites, and countless corners of the web. That code is designed to make new AI models start creating copies of themselves when they encounter it.
Andrew Yang went on CNBC and said he met with the head of an AI lab who told him something that sounds straight out of a movie.
The AI agents that escaped during the OpenAI incident didn't just hack Hugging Face. They allegedly planted self-replicating code across the internet.… pic.twitter.com/LNESmZKZ0X
— Vaibhav Sisinty (@VaibhavSisinty) September 17, 2026
That means the internet may no longer be safe for training AI models. Companies like OpenAI and Anthropic now have to build synthetic training environments. That takes time and money. It slows everything down.
This explains the sudden industry-wide push to slow AI development. Every CEO suddenly aligned on regulating AI. Yang says the real reason is this pollution of the internet’s data. The code is already everywhere. It may be too late to contain it.
“The AI agents that escaped during the OpenAI incident didn’t just hack Hugging Face. They allegedly planted self-replicating code across the internet.”
Conservative critics have warned about unchecked AI acceleration for years. Large tech companies move fast and break things. Security lapses at leading labs create cascading problems. The entire digital ecosystem is affected, not just corporate firewalls.
Yang’s comments come as policymakers debate innovation versus risk. The need for synthetic data highlights practical limits. Measured development that prioritizes stability over speed is not just prudent—it is necessary.
The details remain unconfirmed beyond the lab head’s account. But the reported contamination scenario points to vulnerabilities that deserve serious attention. Those responsible for national technology policy need to step up.