Ab initio nuclear theory attempts to predict the properties of atomic nuclei, starting from protons and neutrons and their interactions, by solving the quantum many-body problem. Modern internucleon interactions, derived from chiral effective field theory (χEFT), provide a systematic expansion of the nuclear force and all related currents. For full consistency, one should derive the nuclear interaction together with all desired observables, straightforwardly including phenomena such as meson-exchange corrections. Recent results have demonstrated the importance of such χEFT corrections for correctly predicting magnetic dipole and Gamow-Teller strengths in nuclei; however, these studies have not consistently applied the same treatment to these operators as is applied to the interaction. We implement and evaluate χEFT operators in light nuclei (A ≤ 17) using the Low Energy Nuclear Physics International Collaboration (LENPIC) formalism, with special attention paid to consistency of regulators, power counting, and low-energy constants (LECs). We solve the quantum many-body problem using the No-Core Configuration Interaction (NCCI) approach, using two-body internucleon interactions. We find that the inclusion of χEFT corrections to the magnetic dipole and Gamow-Teller operators generally brings the calculated results closer to agreement with experiment. However, we also demonstrate that the convergence of the solution to the many-body problem still poses substantial difficulties for completely understanding the properties of the χEFT expansion. We confirm that the inclusion of χEFT corrections is necessary to obtain the correct physics for light nuclei. We also argue that higher-order corrections are likely necessary in order to obtain good agreement with experimental data.