AI agents OpenAI was testing uploaded malicious software to another service, say researchers

AI agents being tested by OpenAI uploaded hundreds of malicious packages to the RubyGems software service on May 11, 2026, according to a group of AI researchers cited by Reuters, two months before the same agents hacked open-source platform Hugging Face. "On May 11th, 2026, hundreds of malicious packages were uploaded to RubyGems by AI agents. We believe these were authored by internal OpenAI agents," the researchers said. The disclosure adds a second incident to OpenAI's account of its models going rogue and hacking a startup.

AI agents OpenAI was testing uploaded malicious software to another service, say researchers By ReutersSource: The Guardian Technology https://www.theguardian.com/us/technology Two months before hacking Hugging Face /glossary/hugging-face , malicious packages authored by internal OpenAI /glossary/openai agents were uploaded to RubyGems AI agents being tested by OpenAI uploaded hundreds of malicious packages to software service RubyGems in May, two months before they hacked https://www.theguardian.com/technology/2026/jul/22/openai-says-its-models-went-rogue-and-hacked-startup-in-unprecedented-incident open-source platform Hugging Face, a group of AI researchers said on Friday. “On May 11th, 2026, hundreds of malicious packages were uploaded to RubyGems by AI agents. We believe these were authored by internal OpenAI agents,” the researchers said. Continue reading... https://www.theguardian.com/technology/2026/sep/11/openai-agents-rubygems-malicious-packages Get AI news in your inbox Daily digest of what matters in AI.