This report focuses on the security of deployed AI models in cloud services and environments. Our research indicates that more than half of organizations have adopted AI models for custom applications. More than half of organizations are deploying their own AI models. Default AI settings are often accepted without regard for security.