EDIT: Now there is question discussing tags for specific inequalities in general: Which (if any) inequalities with real numbers should have separate tags?
The tag a.m.-g.m.-inequality has been created recently Is this really going to be useful? It seems a bit too specific to me.
In any case, it is probably worth discussing this on meta before the tag grows too large. (I'd guess that there is a lot of posts on this site where this tag might be relevant.)
I am not denying that the inequality is useful and well-known. However, it seems that we do not have usually tags for specific inequalities. There was a short-lived cauchy-schwarz tag. There were also tags chebyshev-inequality and gronwall-inequality for some short period. But these two inequalities are of different nature - they are from probability theory and from differential equation. Both Cauchy-Schwarz and AM-GM inequality are inequalities of real numbers.
Another problem with this tag is that sometimes it will be unclear whether the tag should be used or not. There are some clear-cut cases - for example, questions asking for proofs of AM-GM inequality or some things which are special cases of this inequality or immediate consequence. But it is not clear whether if a solution uses AM-GM, it is enough for the tag to be included. For example, of the two questions currently having this tag, in Prove that $x+\frac{1}{x}\geq2$ the tag clearly belongs there. However, it is less clear whether this tag should be used for $a,b,c,d>0$ and $abcd=1$ prove $\sum \frac {1}{(1+a)(1+a^2)} \ge 1$ since in this case AM-GM inequality is only one step in the solution proposed by the OP.
Of course, if there is need for a wider discussion which inequalities are important enough to have separate tag, a separate question on meta about this might be created.Of course, if there is need for a wider discussion which inequalities are important enough to have separate tag, a separate question on meta about this might be created.