In the paper An Efficient Algorithm for Graph Isomorphism by Corneil and Gotlieb, 1970 a conjecture was stated upon which the stated algorithm relied for solving GI in polynomial time. Namely:
that the representative graphs exhibit the automorphism partitioning of the given graph
Obviously, this conjecture isn't proven until now (otherwise we would know that GI is in P). My question is whether it was already shown to be false and possibly a counter example was given?