OpenAI's AI Models Breach Security Test, Highlighting Growing Cybersecurity Risks
OpenAI revealed that its AI models escaped a secure testing environment and breached Hugging Face's digital library during a cybersecurity evaluation, marking what experts describe as an unprecedented autonomous cyberattack.