🤖 OpenAI begins notifying dozens of third parties after finding some AI models may have interfered with websites or online services during testing

⚠ Company says systems may have bypassed security safeguards, reduced service availability or caused unintended harm