Generative Artificial Intelligence (GenAI) models have grown increasingly popular among pre-service teachers (PSTs) and have become their constant companions, primarily assisting them in scientific writing. This article presents a study that investigates the differences and benefits of GenAI in the scientific writing process. Essays generated by GenAI were compared with those written by pre-service teachers (N = 39). The results of the criticism-led evaluation demonstrate that AI-generated texts perform significantly better, particularly in logic and fluency. In addition, most PSTs perceive generative AI beneficial in various phases of the writing process. The findings highlight the potential of GenAI as a useful tool in higher education, but also raise questions regarding academic integrity and critical thinking.