Artificial intelligence is becoming part of everyday work in many fields, and construction is no exception. People use it in different ways, from letting it handle entire tasks to using it as a tool that supports their own judgment. How and when it is introduced into the work process can change the way people rely on it. In construction safety, where mistakes can have serious consequences, we still do not fully understand how the timing of artificial intelligence input shapes decisions. This project looks at what happens when safety professionals see information from artificial intelligence before they make their own assessment, compared with seeing it after they have already formed an opinion. By studying these different sequences, we aim to understand how artificial intelligence can support good safety decisions and where it might create new risks. The goal is to offer clear, practical guidance on how artificial intelligence should, and should not, be used in safety related workflows so that it truly helps people make better informed choices on construction sites.