Navigate assessment design and generative AI

The advent of generative AI (GenAI) is transforming the way we envision, plan, and design assessments. Given this increasingly challenging terrain that includes the intense speed of change, the ever-increasing number of tasks this technology can accomplish, and its unenforceability from an academic integrity perspective, informed decisions that work towards building a toolkit of possibilities as opposed to firm directives may well be the best path forward. 

As part of reflecting on your assessment decisions for your course(s), be sure to review the Guidelines for Teaching with Generative AI to help you determine your overall position on this technology, as well as how you are going to address it with your students.

Important considerations

The recommendations on this page will need to be revisited and revised regularly given the fast-paced evolution of generative AI.

1. As outlined in the Guidelines for teaching with Generative AI, detecting students’ inadmissible use of GenAI is not recommended given that there are no reliable detection tools or methods that can keep pace with this rapidly evolving technology. Detection tools have also been shown to have a high rate of false positives which can lead to students being accused of using GenAI when in fact the work is their own (Weber-Wulff et al., 2023). Evidence also points to a disproportionate number of false positive identifications in neurodivergent students or those for whom English is not their first language (Gegg-Harrison & Quarterman, 2024; Rafiq, Qurat-ul-Ain & Afzal, 2025).
2. In addition to issues of unreliability and inequity in the detection tools, it is also an unethical practice to share students’ intellectual property with large language models (LLMs). At the moment, Concordia has not approved of or acquired the use of any online AI detectors and therefore the use of them by staff or faculty is not permitted due to privacy laws and regulations (see the University’s obligations concerning Privacy & Protection of Personal Information).
3. To learn more about the University’s obligations when acquiring or using new software that could capture and/or share personal information contained in learner work, you may consult the Privacy Impact Assessment and for information about available resources, please consult the IITS Service Catalogue.
2. Instead of trying to detect students’ use of Generative AI, explain the importance of academic integrity and teach the ethics of responsible use. Articulate your position on GenAI clearly within the classroom, on Moodle sites and, most importantly, by means of a syllabus statement. See sample syllabus statements for permitting, prohibiting, or inviting the use of Gen AI on your course outline. Given that individual instructors have different positions on GenAI and messages will vary from course to course, it’s especially important to inform students in writing about your position.
3. Stating when and how GenAI can be used in your course in a syllabus statement and discussing it with your students is an opportunity to be transparent about your position and values and to improve their critical digital literacy. 
3. In addition to the syllabus statement on the use of GenAI in your course generally, include specific allowable uses with each assignment description as these may vary. This assignment template can be adapted for each assignment in your course. Not sure how students are using GenAI? Use this worksheet to test whether you find different uses acceptable/unacceptable or only acceptable in certain conditions.
4. If allowing or requiring the use of GenAI, protect instructor and students’ privacy and ensure secure data management, as well as consensual data collection and processing. Review Concordia’s Privacy and Protection of Personal Information Policies and allow students the option to opt out. Consult the Concordia Educational Technology Guidelines for Faculty and Students before using third party technology tools. At Concordia, all staff and students have access to Copilot Chat through an institutional Microsoft 365 license. Copilot Chat is available on the Edge browser or through the link copilot.cloud.microsoft. It requires authentication (using Concordia credentials) and it offers Enterprise Data Protection: encryption, privacy and access control, copyright protection and the protection against the usage of data to train foundation models.
5. The following icon in the right top corner of the page means Enterprise Data Protection is active:
5. Consider potential biases in GenAI development and output and ensure all students have access (most AI tools are not designed with accessibility in mind).