Sometimes the media picks up on odd ideas from science kit and popularises them prematurely because they seem radical, Years ago it was suggested that Megatherium preyed sometimes on glyptodonts, and almost immediately small ground sloths in the Caribbean were implicated as related to the chupacabras(!). Isotopic studies of Neotropical ground sloths and glyptodonts refute the idea they consumed meat regularly. Though the tree sloths supposedly consume some animal protein so the smaller ground sloths also might have.

Ground sloths are maybe relevant for human origins because they are among mammals with browsing, upright adaptations. Besides any non-leaping bipedal gait being rare among mammals, the gorilla is on a similar browsing path having been likened to chalicotheres and the giant panda. Furthermore though tree sloths are nested within sloths that walked on the ground, most of the smaller generalised species were at least able to climb. How often do arboreal animals adapt to non-climbing esistences? Besides humans I can think of ground hornbills, the patas, and the giant anteater.

The diet of extinct giant Xenarthrans is a debated topic, especially for ground sloths, for which herbivory, insectivory, and carnivory through scavenging or active hunting have been suggested. In this study, stable carbon isotopic composition of collagen and carbonate fraction of well-preserved fossil bones was used as a tracer of trophic level. > 200 modern mammal bones of species with a variety of diets were used as a reference dataset. The good preservation of the carbonate isotopic composition was checked using different diagenetic indicators and by using fossil carnivorous and herbivorous taxa from the same sites as test-taxa. The results on modern mammals indicate a clear distinction in the difference between the carbon isotopic composition of carbonate and collagen between faunivores (carnivores, insectivores, piscivores, omnivores) and herbivores. Using this framework, the results on the fossil megafaunal Xenarthrans indicate a herbivorous diet for both Glyptodontids and are also consistent with a herbivorous diet for giant ground sloths, including Megatherium. The hypothesis that Megatherium could be a cryptic meat-eater or an insectivore is not supported by the results obtained in the present study.
http://www.sciencedirect.com/science...42937X16304944