Many scientists are working on algorithms and models to learn from COVID-19 data. Some useful data sets are available at Kaggle The University of Montreal has a started database of images from CT scans and chest X-rays from actual patients. Finally, for a few months now, Johns Hopkins has a great dashboard using ArcGis that displays the growth of the virus. The dataset from Johns Hopkins is available at this GitHub site.
Finally, here is a preliminary paper that attempts to create a rapid AI model to detect the disease and also track the impact of the effects on the lungs.