Model collapse is a risk for enterprise ai that may be already happening. The phenomenon is linked to the use of synthetic data and ai-generated content, which can potentially cause ai models to fail. However, researchers may have found ways to prevent or stop model collapse, including the use of a single real-world data point.