The divergence of Homo from gracile australopiths has been described as a trend of decreasing dentognathic size and robusticity, precipitated by stone tool use and/or a shift to softer foods, including meat. Yet, mechanical evidence supporting this narrative is sparse, and isotopic and archaeological data have led to the suggestion that a shift away from a gracile australopith-like diet would not have occurred in the most basal members of Homo but rather only with the appearance of Homo erectus, implying that the origin of our genus is not rooted in dietary change. Here, we provide mechanical evidence that Homo habilis exhibits an australopith-like pattern of facial strain during biting but, unlike most australopiths, was not suited for a diet that required forceful processing by the molar teeth. Homo habilis was at elevated risk of distractive jaw joint forces during those bites, constraining muscle recruitment so as to avoid generating uncomfortable/dangerous levels of tension in the joint. Modern humans have similar limitations. This suggests that selection on skeletal traits favouring forceful postcanine processing was relaxed by the earliest stages in the evolution of our genus, implying that dietary or food processing changes played an important role in the emergence of Homo .