<p>This paper studies the statistical inference of the inverted exponentiated Rayleigh distribution under generalized adaptive progressive hybrid censored competing risk data. The likelihood function is obtained by considering two risk factors of failure. Further, maximum likelihood estimates are obtained as a point estimation. In addition, the observed Fisher information matrix is calculated, and the asymptotic confidence intervals are obtained in the sequel. In the Bayesian framework, the importance sampling method is employed under the squared error loss function, with the a priori chosen model being gamma. The highest posterior density intervals are also obtained in Bayesian estimation. To evaluate the performance of the proposed methods, a simulation study is conducted using R software, and numerical comparisons of the estimators have been discussed. Two failure data sets are analyzed for the applicability of the proposed model in practice. Finally, four optimality criteria (A, D, F, V) are chosen to obtain optimal designs.</p>