Shilin and Suslov classified Arstanosaurus as a hadrosaurid in 1982, but by 1995 Nessov was convinced it was a neoceratopsian, due mainly to a similarly double-rooted tooth. Later still, Norman and Kurzanov announced that it had been incorrectly described in the first place; it didn't have double-rooted teeth at all, and the rest of its remains sadly lack diagnostic characters, making it a nomen dubium. Nevertheless, they tentatively assigned it to Lambeosaurinae based on the shape of its upper jaw, lamented a lack of features in common with any other Kazakh hadrosaurid (mainly because there are no like-for-like parts to compare), and pinpointed features in common with the Chinese Bactrosaurus.
A juvenile skeleton that was previously (and unofficially) tagged "Gadolosaurus" by Tsunemasa Saito whilst it toured Japan as part of a Soviet exhibition was recently sunk into Arstanosaurus by David Norman with no attempt to justify the move. It, and Arstanosaurus in its entirety, in fact, may belong to Batyrosaurus which was found in exactly the same area, though the only comparable bits are teeth... and they're different.
The species epithet, akkurganensis, means "from Akkurgan" in Latin.