Adversarial Testing

OpenAI Releases Open-Weight LLMs With Safety Challenge and Azure AI Foundry Integration

A $500,000 red teaming contest will assess the models’ robustness under Apache 2.0 licensing before widespread rollouts