In the second quantization language, stimulated emission corresponds to the bosonic property
$$
a^\dagger|n\rangle=\sqrt{n}|n+1\rangle.
$$
That is, the more quanta a mode contains, the higher is the probability of emission into this mode (the relevant coupling term containing operators $a, a^\dagger$ - see this discussion for an example of light coupling to a two-level atom, and this one for coupling in a semiconductor).
On the other hand, for Fermions
$$
c^\dagger|n\rangle = \begin{cases} 0, \text{ if } n=1,\\ |1\rangle, \text{ if } n=0\end{cases}
$$
That is, if a mode already contains a fermion, another fermion cannot be added to this mode (i.e., its addition is "inhibited") - which is just the Pauli exclusion principle.