Yes, it's just a linear algebra stuff. Probably, it's more convenient to look at  the Weil pairing between the  $\ell$-adic Tate modules of $A$ and its dual, taking into account that these modules are free $O\otimes Z_{\ell}$-modules of the same rank.