Probably a WS loanword (with Goetze 1947:246‒247). This hypothesis is attractive in view of the fact that Akk.
emu was clearly polysemic (“fatherin-law”/“son-in-law”) in OB and OA, whereas ḫatanu is only sparsely attested in early periods (Kogan 2014:104‒105). Also the morphological shape of the Akkadian word (non-syncopated *a in the second syllable) is unusual.