It seems to me that what happens depends on the conductivity of the metal.
A metal is a lattice of ions swimming in a sea of electrons, so when you
impose a microwave (electric vector in rotary motion) the photon-electron
interactivity coefficient (frequency dependent) the electron will be put
into motion (ie, electric current). Given the resistivity (reciprocal of
the conductivity) there will be an I-squared R loss. This energy term via
the metals heat capacity would generate some temperature rise. I do not
have the numeric databases available so can't estimate the temperature
increase.

On the other hand, I once accidently put a ceramic dinner plate that had a
decorative gold rim into the microwave. The sparking was impressive and
the grub on the plate never warmed up at all. I expect the mobility of the
gold rims electrons sucked up all the microwave energy. Someone else told
me that was a good way to burn out a microwave.

Do not induction furnaces operate at much lower frequencies.