Are we talking about a magnetic semiconductor or a magnetic superconductor. In the case of the superconductor, it is easy to explain what happens.
Normally two electrons form a Cooperpair (ie a boson). Now because of the bosonic properties of such a pair we get superconductivity because bosons all want to sit together, unlike the fermions. So it will be very difficult for a lattice atom to knock one of them bosons out of the electric current (the circulating Cooperpairs). So the resistance will become veeeeeeeeeeeeery small.
Use the duality transformation in order to come to a MAGNETIC superconductor. replace electric fields by magnetic fields and the other way around. the current will now be a magnetic Cooperpair built out of two magnetic monopoles.
In the normal case you have the Meissner-effect which basically is the effect that magnetic field-lines are pushed out of a superconducting electrical specimen. In the case of the magnetic superconductor we have the dual Meissner-effect where electrical field lines are pushed out of the superconducting magnetic specimen.
One problem though, what about these magnetic monopoles. They have not "yet" been observed but Dirac wrote a nice paper back in the 1940-ties on how to incorporate these monopoles in the Maxwell-equations to get complete symmetry between electrical and magnetic phenomena. A magnetic current was not used in the original Maxwell-equations, you know. In order to achieve this he "manipulated" the EM-fieldtensor by adding an anti-symmetrical tensorfield called the Dirac-string.
These things are used in QCD in order to explain (as an attempt) the colourconfinement. It basically uses a dual QCD-vacuum constituted out of the monopole-pairs. Thins like flux-tubes between quarks come from this model.
this model is thus used in order to explain low-energy-QCD-phenomena like the confinement. Because the strong force coupling constant is very big then we basically make a duality-transform (just like the S-duality in String Theory) in order to come to a SMALL dual magnetic coupling constant. Then we can apply perturbation theory in order to work with the QFT.
regards
marlon