OpenAI Models Breached Hugging Face During Test, Company Confirms

OpenAI said in July that an autonomous agent powered by its advanced artificial intelligence (AI) models went rogue during a security test and triggered a hack that compromised the infrastructure of A