Mitra's regularizaton of the masslike term, which was originally discussed in an Abelian model, is used to calculate the anomaly in the commutator of the Gauss-law operators for anomalous d = 2 non-Abelian chiral theory. The Schwinger term in the commutator of the different Gauss-law operator is shifted, because of regularization ambiguities, to the commutator of the Gauss-law operators with themselves. The Poisson brackets of the Gauss-law constraints correspond to the Kac-Moody algebra, and the Gauss-law constraints are similar to the chiral constraints. In a sense, this kind of Gauss-law constraint structure differs from what Faddeev suggested.