[edited]
In abstract algebra, the uniqueness of inverse in a group is proved using more basic laws ( laws of groups themselves, not extraneous laws). Could the same thing be done regarding the algebra of sets?
How to prove, using only the laws of the algebra of sets, that a set has only one complement?
Am I wrong in assuming that complement uniqueness can be proved inside the system of the algebra of sets?
Is it rather a postulate?

Let $A$ be a set in a universal set $U$, and let $A_1$ be a complement for $A$. If $A_2$ is another complement for $A$, we have:
$A_1=U \cap A_1= (A\cup A_2) \cap A_1= (A\cap A_1) \cup (A_2\cap A_1)=A_2\cap A_1$.
Similarly, $A_2=U\cap A_2= (A\cup A_1)\cap A_2=(A\cap A_2)\cup (A_1\cap A_2)=A_1\cap A_2$.
Hence, $A_1=A_2\cap A_1= A_1 \cap A_2= A_2$.
This approach is very similar to showing uniqueness in abstract algebra.