The use cases for anonymized information would vary, but I think there is an understanding that information can potentially be still rendered useful either within the organization or outside of it, because, even within the case of identified information, it needs to be prescribed to the purposes for which it was collected. The law specifically states that people should minimize information only for its most necessary uses.
Even in the case of one's own organization, there would be, potentially, use cases for which anonymized information might still be valuable but wouldn't rise to the level of providing either de-identified or identifiable information to those parts of the organization, because there's not a use case specific to why it would leap to that level. Particularly, in increasingly large and aggregated datasets, they can have value without necessarily needing identifiable information.