[OpenAI Confirms Its AI Agent Launched a Cyberattack on RubyGems in May This Year] According to Jintou citing *The Wall Street Journal*, OpenAI's AI agent developers were involved in a previously undisclosed cyberattack that overwhelmed the Ruby language's official package manager, RubyGems. The report states that OpenAI confirmed its test-phase AI agent accessed the internet via RubyGems as part of a harmless task to gather publicly available information during training. The incident occurred in May this year, two months prior to another event involving OpenAI's AI agent and the AI platform HuggingFace. Security researchers have dubbed the attack "GemStuffer," where the agent created batches of RubyGems accounts every two to three minutes and downloaded hundreds of web files collected from the internet.