Universal quantum (semi)groups and Hopf envelopes: Erratum
In \cite{F} there is a statement generalizing the results in \cite{FG}. Unfortunately there is a mistake in a computation that affects the main result. I don't know if the main result in \cite{F} is true or not, but I propose an alternative statement (Theorem 3.7) that was actually the main motivation in \cite{FG}. This statement answers in an affirmative way the question whether the localization of the FRT construction with respect to a quantum determinant is a Hopf algebra, in case the Nichols algebra associated to the braiding is finite dimensional.