With the advent of Generative AI (GenAI), academic institutions are facing a pressing need to adapt assessment strategies to ensure that assessments genuinely reflectstudents’ own work, rather than the output of AI tools. A promising approach to address this challenge is frequent, small-scale, in-class assessments. However, this approach significantly increases the assessment workload, quickly making it unmanageable. To mitigate this, our project proposes leveraging AI itself to assist in evaluating some of these assessments. Specifically, we aim to investigate the effectiveness of AI in grading a selection of in-class assessments across three distinct academic disciplines. The AI’s grading will be compared against human evaluations using the same assessment instrument to assess its accuracy and reliability. Furthermore, to ensure that students’ work is not used in training commercial AI models like GPT-4, without student consent, we plan to deploy an instance of an open-source Large Language Model (LLM), such as Llama2. The project will develop a simple web application for students and academics which allows learners to submit pieces of writing, and the teacher to review and compare their review with the AI review, and then distribute the feedback. Students will be co-designers of this system using a focus group and survey to provide feedback on early prototypes as they are piloted in across a small number of modules. This project aligns with all three SALTE themes: 1. Leveraging technology for immediate, personalized feedback, it adheres to Universal Design for Learning (UDL) principles, benefiting all students by offering timely, tailored support regardless of learning style or ability. 2. It aligns with “Next Steps for Teaching and Learning: Moving Forward Together” by integrating digital tools in assessments, embracing a blended learning approach with AI technology. Our use of an open-source Large Language Model for assessment upholds digital and open teaching policies, ensuring ethical and open-source digital transformation. 3. Byimplementing AI in assessment workflows, this project engages with Generative AI challenges and contributes to national resources. It supports the National Academic Integrity Network (NAIN), providing tools to uphold integrity in evolving educational environments.