This paper presents a preliminary case study on detecting long-form, in-the-wild audio deepfakes. It investigates challenges in real-world spoofing detection, evaluates robustness under diverse acoustic conditions, and explores biometric signal analysis methods to improve detection reliability.