The maximum will be $n+1$.
If there were more than $n+1$ sets then for any specific set $A$ there will be at least $n +1$ pairs of sets where $A$ is one of them. $A$ will need a distinct element for each of those pairs so $A$ will have at least $n+1$ distinct elements violating condition $1$.
So the maximum is $\le n+1$.
$n+1$ is possible. There will $n$ pairs for $A$. Those can be the disinct elements of $A$.
If you need a concrete example: Let each $A_i$ be a set of unordered pairs. I.e. sets of two numbers where order doesn't matter so $\langle a, b\rangle$ is considered to be the same as $\langle b, a\rangle$
Let $A_i = \{ \langle i, j\rangle| j \le n+1; j\ne i\}$. i.e. the set of all pairs containing $i$ (pairs of distinct numbers within a range of $1$ to $n+1$ of with $i$ is one of the pair).
Then $A_i \cap A_j = \{ \langle i, j\rangle\}$ but $\{\langle i, j\rangle\}$ will not be in any other $A_k$. (as $i \ne k, j\ne k$).
If that's too vague:
Let $p_i =$ the $i$-th prime.
Let $A_k = \{p_i*p_k|i \le n+1, i \ne k\}$.
Then each $A_k$ will have $n$ composite numbers as elements. If $i \ne j$ then $A_k \cap A_j = \{p_k*p_i\}$. This number $p_k*p_i$ will not be in any other set than those two as $p_kp_i$ has no other prime factors. So all conditions are met.