Sufficiency is joint: $\bar X$ alone is not sufficient for $\mu$ when $\sigma^2$ is unknown

Let $X_1,X_2,\dots,X_n$ be a random sample from $N(\mu,\sigma^2)$, where $\mu$ and $\sigma^2$ are unknown. Then which of the following statements is not correct?