Understanding the influence of clouds on amplified Arctic surface warming remains an important unsolved research problem. Here, this cloud influence is directly quantified by disabling cloud radiative feedbacks or “cloud locking” within a state‐of‐the‐art and well‐documented model. Through comparison of idealized greenhouse warming experiments with and without cloud locking, the influence of Arctic and global cloud feedbacks is assessed. Global cloud feedbacks increase both global and Arctic warming by around 25%. In contrast, disabling Arctic cloud feedbacks has a negligible influence on both Arctic and global surface warming. Interestingly, the sum of noncloud radiative feedbacks does not change with either global or Arctic‐only cloud locking. Notably, the influence of Arctic cloud feedbacks is likely underestimated, because, like many models, the model used here underestimates high‐latitude supercooled cloud liquid. More broadly, this work demonstrates the value of regional and global cloud locking in a well‐characterized model.