With the booming rise in GenAI technologies, specifically powerful Language Models (LLMs), such as ChatGPT are going mainstream across areas from music and creative writing to marketing—the classroom setting intensifies several issues with their adoption in education. Originally heralded as an education aid, students instead turn to ChatGPT for crafted text of such startling high-quality school assignment submissions sometimes too closely resembling random writing from a human rather than genuine original work. The research presented addresses the urgent problem of AI-based facsimile with a proposal for an original method to differentiate whether specific text was written by or created through artificial intelligence in student assignments. In order to answer that conflict, we developed a faculty assisting system which uses a fine-tuned BERT-based Disguised Language Transformer use case for educational content to quantify how influenced a students’ submitted work is by way of AI. Moreover, we use the T5 model to produce targeted questions from student responses to confirm comprehension of the syllabus through the eyes of the teacher. This is deployed through a web portal that allows teachers to upload assignments and be given granular analytics on how much text was AI-authored versus authored by humans, along with auto-suggested questions for more evaluation data. We enable educators to detect degrees of AI involvement and generate contextual questions, providing a powerful tool for discouraging cheating while promoting authentic learning.