Artificial intelligence (AI) is a major player in the drug development process. Artificial neural networks, such as recurrent or deep neural networks, are the main forces behind this research. Applications for forecasting attributes or behaviors, such ADMET and physicochemical properties have increased recently. The technology's strength in these sectors is reinforced by quantitative structure-property relationships (QSPR) or quantitative structure-activity relationships (QSAR). Artificial intelligence in de novo design guides the synthesis of novel, functional, biologically active molecules toward desired properties. Numerous examples show how successful artificial intelligence is in this field. Drug discovery can be integrated with synthesis planning and ease of synthesis, and it is predicted that in the near future, computers will be used for an increasing amount of automated drug discovery. The pharmaceutical industry is currently having trouble keeping up its drug development programs because of increased R&D costs and declining productivity. In this paper, we investigate the main factors influencing the attrition rates in the approval of new medications, possible approaches and kinds of AI-based software to enhance the effectiveness of the drug research process, and cooperation between the titans of the pharmaceutical industry and AI-powered drug discovery companies.