<p>Serverless computing is a cutting-edge trend in cloud computing that offers a platform for creating event-based, flexible functions in a scalable setting. Although this approach presents a promising development model, it has also introduced a novel type of cyber-attack called Denial of Wallet (forced financial exhaustion). The threat of denial of payment and its potential attack vectors have been defined and identified in this work. We also show how this new form of attack could, in fact, evade the current mitigation systems designed to respond to a similar style of attack, denial of service. Our goal is twofold; First, a brief and useful overview of this new attack paradigm will be provided. Secondly, we simulate these attacks in an isolated environment and as our primary contribution, we apply and evaluate a Machine Learning (ML) based approach for DoW detection. We use a decision tree model to establish a foundational performance baseline for this emerging threat and compare its effectiveness against both a traditional rate-limiting defense and a system with no protection. While rate-limiting proved most effective in our flood-attack scenario, the proposed ML model also significantly reduces the financial impact of the attack on the deployed function.</p>