OpenAI has imposed stricter controls on its upcoming Astra AI model due to preliminary tests indicating potential for advanced cyber capabilities. The company is enhancing safeguards as it continues to evaluate the model's risk profile.
