Key Principles

High-level principles:

The University recognises that AI applications have significant positive uses to enhance inclusive learning and facilitate reasonable adjustments for students with disabilities. 

1. Critically engaged generative AI literacy is a core contemporary skill. All students should be encouraged to critically engage with generative AI outputs and acknowledge the capabilities, limitations and costs of its use.
2. Staff should actively engage students in discussions around generative AI literacy. It is important for students to be given the space to speak about generative AI outputs, to understand how types of generative AI, such as large language models, work, what they do, and their drawbacks and limitations.
3. It is expected that students develop critical literacy with respect to generative AI content and that all outputs are seen as potential tools rather than definitive authoritative sources. 
4. Staff and students should be cognisant of the ethical implications of its use and actively seek to remove and challenge inherent bias and potential stereotypes from generative AI outputs.
5. All users should be aware of the significant environmental impacts of the use of any computational resources. 
6. The University recognises that AI is impacted by fairness of access, acknowledging that advanced AI functionality often lies behind a paywall. To ensure that all students have access to approved generative AI resources the University has approved the use of Microsoft Copilot for all staff and students. This has been chosen because of its enhanced data security and like other Microsoft programmes it is private to the user. University prescribed generative AI platforms, currently the University’s version of Microsoft Copilot, should be the only generative AI platform used for study and assessment, where permitted.
7. Privacy and fairness of access are of paramount importance when using AI tools.   
   1. Students and staff should not upload sensitive, personal, or confidential information to generative AI platforms.
   1. University data (including unpublished research or student/staff information) must not be uploaded to open generative AI platforms.
8. No assessment should be designed that implicitly requires access to premium AI tools which are not freely available to students.