What you describe seems not to be detection of spam, but rather use of copied content with improper attribution.
It is encouraged to quote "verbatim" the most relevant part of the material to be found by following a link, where attribution is provided and the quoting does not (by its extensiveness and lack of authorization) infringe on copyright or constitute plagiarism. Alternatively the OP may prefer to restate the ideas in summary form using their own wording. See here under Provide context for links.
If the (absence of) markup blurs the line between copied and original exposition (perhaps by omitting any original exposition), then this could be a problem with improper attribution.
As a first step I'd probably leave a Comment for the OP, suggesting they highlight what material is copied and thus distinguish it from original exposition. E.g. "Please edit your otherwise useful Answer to delineate between the words you are contributing and those you've found in the linked content."
If that does not result in suitable improvement, and the difficulty in supplying an appropriate "friendly" edit is significant, then it would be worth flagging for moderator attention (IMHO).
I had the occasion today to revise an Answer posted in 2012 on sister site SciComp.SE to supply better indication (for casual Readers) of the content to be found by following the link I gave.
On the other hand when I review items in the queue for content, this typically includes assessing the validity of the links provided (so that spam or unacknowledged self-promotion can be detected). If this causes me to fail the review audit, it seems worth the "price".