Abstract
With the rise of generative artificial intelligence, previous barriers to producing written content in the form of papers and grants are decreasing, resulting in a rapid rise in the level of submitted material. This opinion piece first describes a number of already existing structural problems within current scientific research practice that relate to publication output, managerial science styles, and the operation of national granting systems that are especially susceptible to further exploitation with the use of an artificial intelligence-based writing assistant. It then proceeds to offer a number of recommendations that may help to fix these problems, thereby creating better working environments and higher quality research output. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s12551-025-01324-8.