Let $R$ be a commutative ring and $A \subset R$ be any ideal. The ideal $A$ is finitely generated if every $a \in A$ can be written in the form $a = r_1a_1+r_2a_2+ \cdot + r_na_n$ where $r_i \in R$ and we write $A = \langle a_1, a_2, \cdots , a_n\rangle$.
Is it correct to conclude that any element of the form $r_i^ma_i^n \in A?$
My reasoning: Since any $a\in A$ can be written as $ a = r_1a_1+r_2a_2+ \cdots + r_na_n$,
we get $a_i \in A \implies a_i^2 \in A \implies a_i^n \in A \implies r_ia_i^n \in A \implies r_i^ma_i^n \in A$
Your definition is not written correctly. In your definition, the elements $a_1,\ldots,a_n$ are not specified and so are free in your sentence. That means that either you are saying
or else you allow the $a_i$ to depend on $A$. In the first case, very few ideals would meet the condition; in the latter, all ideals would meet the condition.
The correct way to phrase it is to specify the $a_i$ first. So it should read:
But note that what you are asking about has nothing to do with whether the ideal is finitely generated. Simply:
If $n=1$, this follows because $A$ is an ideal, so $a\in A$, $r^m\in R$ implies $r^ma\in A$. For $n\gt 1$, you can let $s=r^ma^{n-1}\in R$, and then $a\in A$, $s\in R$ implies $sa = r^ma^{n-1}a = r^ma^n\in A$.
Nothing to do with finite generation, everything to do with being a ring and an ideal.