Fraud Detection is a major concern these days because of digitalization. We are totally dependent on online transactions these days for even very small needs. There is no doubt that online transactions have made our life very easy but it has increased risk on other hand. And this risk can be very harmful one day. Confidential data is being stolen by the different apps and it is sold in international market. Which later on comes to us in totally different and very harmful way. So why not to use technology again to stop these risks and flaws. Various ML techniques has been observed by researchers but Auto ML is yet not discovered on a wider platform. Therefore, this paper at first aims to explore the trending technology Auto ML. Then a model for evaluating Auto ML is suggested and analysed with different classification algorithms. The experimental results ascertained the accuracy of Auto ML followed by a comparative analysis of ML and Auto ML.