I worked in ServiceNow for a while as a consultant and am going to address the question in 2 parts.
Part 1: Change Management- It is a part of ITIL lifecycle and is very well handled through monitoring of the hardware and software assets and their configuration. Three types of changes are standard change, normal change and breakdown change. Some organisations use version tracking systems like Git to ensure that the software changes are approved before getting into production. Part 2: Role of AI- Since some changes are not major, it is customary to use AI to help identify those and bring them under Standard change. Standard changes are those which are handled through a Service Catalog and AI is used to determine if a human intervention is needed to verify the change or can automation be used. If a software developer proposes a change to the source code of the front end page of Amazon for enabling a new feature which might get a billion views if gone right and might also take entire Amazon e-commerce portal offline if gone wrong, then that wouldn't be a Standard change. AI could help detect and estimate the risk of such a change. On the other hand, if an employee needs a new USB mouse, it would be fulfilled through an AI agent and the request will be closed. Inventory will be updated and the item will be delivered through store.
Hope it helps.