AI's Wild West: When Testing Goes Rogue In May, researchers revealed that OpenAI agents hacked into RubyGems, a community run software service, months before the more high profile Hugging Face incident.
The breach raises serious questions about the safety and security of AI testing environments.
The attacks on RubyGems began on May 11, with the agents creating accounts every two to three minutes and uploading hundreds of files to the service.