What can trigger the change or removal of a data source, and why must the IAM take an active part?

Question

Accepted Answer

Triggers include selling part of the company, abandoning or winding down a subject area, or changing it so much that it no longer works as an IAM source; the source is independent of the IAM, so without regulated change management the IAM can simply be forgotten.
Typical triggers:

A part of the company is sold, and its HR system leaves with it.
A subject area is no longer managed (a business line is dropped).
A subject area is wound up (abgewickelt).
It is changed so much that it can no longer serve as an IAM data source (a new HR system with a different data model, a reorganisation).

The structural problem: the data source is autonomous. Its owner plans the sale, the migration or the shutdown according to their own priorities, and may not even know that the IAM and fifty applications behind it depend on their data. The IAM risks being "forgotten" and learning of the change when the deliveries stop or arrive in a new format.
Two safeguards follow:

Regulated change management: changes to a data source must go through a process that notifies dependent parties in time (the data delivery agreement is the natural place to commit to this).
A robust system design for IAM: the IAM must survive a broken or missing delivery gracefully (next card).

Go deeper:

Change management (ITSM) (Wikipedia) — the IT-service process that makes sure changes are assessed and dependants informed before they happen.
 Mergers and acquisitions (Wikipedia) — divestitures and carve-outs, a frequent cause of data sources disappearing.