Credit card fraud is a significant challenge in the financial industry, leading to substantial financial losses and eroding consumer confidence. Traditional fraud detection systems struggle to keep up with the dynamic and constantly evolving nature of fraudulent activities, requiring more adaptive and efficient solutions. This study explores the use of the Isolation Forest, an unsupervised machine learning algorithm, to detect anomalies in credit card transactions. The traditional method, Isolation Forest isolates anomalies by identifying points that are few and different from the majority, allowing it to effectively detect potential fraudulent transactions. The paper outlines the methodology of Isolation Forest, detailing how it isolates outliers and flags suspicious activities without the need for labeled data, making it particularly suitable for large-scale, high-dimensional datasets typical in credit card transaction records. Additionally, the application of the Isolation Forest model in a real-world credit card fraud detection context is demonstrated, comparing its performance against conventional techniques. Results show that Isolation Forest not only improves the accuracy of fraud detection but also significantly reduces false positives. This approach presents a promising alternative to conventional fraud detection methods, offering enhanced adaptability and efficiency in addressing the complexities of modern fraudulent behaviours.