All knowledge should be processed by internal rules. Clear criteria specify, how the content is classified and thereby which retention period it belongs to.
Clearly the organisation should filter and prioritise knowledge already at this early stage:
Copies or duplicates develop an uncontrolled own „life“ and are subject to unintended or unauthorised changes. Managing those consumes additional effort by the organisation and may mislead it.
Less content minimises the sources of error and therefore costs.
Typical sources of errors are are E-Mail attachments, non-systematic backups or local/cloud copies of files.