A planetary microlensing signal is generally characterized by a short-term perturbation to the standard single lensing light curve. A subset of binary-source events can produce perturbations that mimic planetary signals, thereby introducing an ambiguity between the planetary and binary-source interpretations. In this paper, we present analysis of the microlensing event MOA-2012-BLG-486, for which the light curve exhibits a short-lived perturbation. Routine modeling not considering data taken in different passbands yields a best-fit planetary model that is slightly preferred over the best-fit binary-source model. However, when allowed for a change in the color during the perturbation, we find that the binary-source model yields a significantly better fit and thus the degeneracy is clearly resolved. This event not only signifies the importance of considering various interpretations of short-term anomalies, but also demonstrates the importance of multi-band data for checking the possibility of false-positive planetary signals.