Researchers have been caught publishing results from cracked Gaussian versions, leading to retractions, institutional investigations, and career damage.