Recent research suggests that diplodocoids disappeared from Patagonia by the Coniacian, replaced by derived titanosaurs. In light of this, and research by Curry Rogers indicating that Nemegtosaurus is a titanosaur, despite diplodocoid-like skull features, isn't it most likely that Antarctosaurus includes material from a diplodocoid-like titanosaur, rather than a true diplodocoid?
At least one diplodocoid (Sciutto and Martinez, 1994) survived in Argentina past the Coniacian. This diplodocid is reported from the Maastrichtian and is said to have elongated neural spines. Tom Holtz originally mentioned this taxon back in 1995 on the list, so perhaps he knows more (like the papers citation).
In any case, Antarctosaurus' cranial remains share derived characters with Nigersaurus, which allies it with rebbachisaurids rather strongly.