Reports on students’ declining writing skills have been a recurring topic in the Flemish media. A trend that will only become more prominent with the recent advancements of generative AI, with ChatGPT as its most famous representative. While this technology can assist during writing tasks, it lacks the core skills necessary for qualitative writing, such as creativity and critical thinking, and therefore cannot replace human writing skills. Research shows that these skills benefit from detailed corrective feedback. Even in ideal circumstances, this task is labour-intensive for teachers, but it becomes even more challenging when considering the growing teacher shortage in Flanders. This project aims to provide a possible solution by developing an AI-based Dutch automated writing support system. The envisioned system will not only be able to correct low-level writing problems, such as spelling and grammar errors, but also to provide feedback on more complex writing issues on the text level. In order to achieve this, I will (i) improve upon existing writing support systems to detect and correct low-level errors, (ii) explore novel AI methods to enable the evaluation of text-level errors and (iii) exploit generative AI to give more meaningful feedback. The core aim of this project is to end up with a writing support system which is able to not only detect and correct writing errors at various text levels but also to motivate writers with the right feedback at the right time.