Arctic amplification and sea ice loss are striking features of recent climate change and are projected to continue. A common approach to assessing the effect of Arctic amplification and sea ice loss on global climate is to perform coupled ocean-atmosphere climate model simulations in which sea ice is perturbed to a future state, in the absence of anthropogenic greenhouse gases. A recent study (England et al., 2022) criticizes this approach, pointing out that sea ice perturbation experiments add artificial heat to the climate system. This paper shows that in a simple energy balance model, sea ice perturbation simulations overestimate the warming caused by sea ice loss by a factor of 1.5-2.  In this talk, I will establish the robustness of artificial warming to model complexity and discuss how sea ice perturbation experiments should be interpreted in light of their artificialness. I will propose a method for accounting for the artificial heat and use this method to demonstrate how these simulations have been misinterpreted in the past.