A researcher assumes a Poisson model $Y \sim \text{Poisson}(\lambda)$ for count data. However, the data exhibits overdispersion ($Var(Y) > E[Y]$). If they proceed with standard Poisson Maximum Likelihood Estimation, how is the consistency of $\hat{\lambda}$ and the validity of the standard errors affected?