From a more general point of view, electric potential is not a scalar, but a component of a 4-vector (http://en.wikipedia.org/wiki/Electromagnetic_four-potential ) - it is not invariant with respect to boosts.
Edit: in answer to question:
Well the potential is just one component of the four-vector. Not the four-vector itself. So isn't it a scalar after all?
As user Ben Crowell says:
No, the way physicists really think about this is that categories like vector and scalar are defined in terms of their transformation properties. A scalar is something that doesn't change at all under any smooth change of coordinates, e.g., a Lorentz boost.
but also we could say that electric potential is a scalar under the rotation group and more generally under the Galilean Group but only a component of a 4-vector under the Lorentz group.
