The project examined the threats and opportunities presented by large language models in relation to assessment integrity. It identified modules and assessment approaches most vulnerable to AI-assisted academic misconduct while exploring ways to strengthen assessment design and teach students to use these tools appropriately. Key activities included an audit and traffic-light analysis of module assessment strategies, supported by case studies demonstrating potential misuse of large language models. A student hackathon generated creative approaches to AI-resilient assessment and explored the educational and employability benefits of incorporating AI into technical teaching. The project also piloted AI-aware teaching and assessment practices, including approaches to online assessment. Resulting case studies and recommendations supported the development of more robust, adaptable, and forward-looking assessment strategies.