I have tried to prove this statement the following way:
We will call $d_1=\gcd(a,b)$, $d_2=\gcd(a,c)$, $d=\gcd(d_1,d_2)$ and $D=\gcd(a,b,c)$.
We know that:
- $d_1 \mid a$ and $d_1 \mid a$
- $d_2 \mid a$ and $d_2 \mid c$
- $d \mid d_1$ and $d \mid d_2$
So, if $d \mid d_1$ and $d_1 \mid a$, we know that $d \mid a$.
The same can be said to conclude that $d \mid b$ and $d \mid c$.
We know that if $D=\gcd(a,b,c)$, this means that $D \mid a$, $D \mid b$ and $D \mid c$.
As we can see both $D$ and $d$ divide $a$, $b$ and $c$, so we can conclude that $\gcd(a,b,c) = \gcd(\gcd(a,b),\gcd(a,c))$.
Is this proof conclusion correct?
You have proved that $d$ and $D$ are common divisors - but are they the same?
$D$ is the greatest so $d\le D$,but you need to prove the reverse inequality to conclude your proof.
For example both $2$ and $6$ divide the numbers $6, 30, 72$ but $2\neq 6$ so you can't include that two common divisors are equal.
It should be easy enough to fill the gap - at the moment you haven't quite tied down the relationship between $d$ and $D$.