With the advent of Google’s Alphafold2 (3D models) and OpenAI’s ChatGPT4 (Large language models), researchers anticipate game changes in drug development. Alphafold2 would allow structure based drug design in every disease area and ChatGPT could help in diagnoses and personalize medicine. Furthermore AI is used to help set up large international phase 3 trials. However, training the models well depends on the quality of the data used and access to positive as well as negative data, disease as well as healthy data. What is the current status?