Number of people die from flooding each year, both in India and outside. Numerous factors can cause floods, such as excessive rainfall, tsunamis, or the failure of infrastructure used to hold water (such as dams, levels, and retention ponds). One can categorize a flood as either aperiodic or periodic. Periodic floods happen on rivers, while aperiodic floods are caused by heavy rain or water logging during wet seasons. Despite the fact that there are numerous methods for detecting floods, the findings are insufficient because the number of flood-related deaths rises year. Therefore, the development of a suitable technique for flood detection is necessary. Techniques for flood detection based on artificial intelligence (AI) and Internet of Things (IoT) are compared to meet the need. The three main technologies that are typically used to detect flooding are IoT, WSN, and MANET. IoT is the most effective technology among these three, with the ability to communicate and analyze issues. Numerous devices are used in this technique to sense. The precision of sensed data will also depend on where the gadget is installed. This study suggests an AI-based IoT model for early flood detection and risk reduction by warning residents living downstream of the river dam. To predict accurate values, the output data of IoT devices is fed to AI algorithms like Linear Regression, artificial neural networks, and support vector machine. ANN has a lower error rate than the other two.