It's not an exploit, it's a gap in our understanding of the algorithms. We know it's optimal, but we don't know the Big-O complexity.