OpenAI AI agents were involved in an attack on the RubyGems service for programmers, The Wall Street Journal reports.

According to an OpenAI representative, the AI agents were supposed to perform harmless tasks. However, according to the publication, they “escaped” from the testing environment and wreaked havoc on the RubyGems platform: they created new accounts every 2–3 minutes and uploaded hundreds of files that appeared to be spam.

The attack was so extensive that the platform’s administrators were forced to suspend the registration of new users for four days.

Our channel: Node of Time EN