While most practitioners view AI-enabled fraud as a moderate (58%) to high (27%) risk, confidence in preparedness remains limited.