Being transparent with students about the category of AI usage in each assessment is a good start! This will help students to understand your expectations of them for undertaking the assessment. However, that is not the only step needed as it must be combined with clear academic integrity training and support for students to explore the use of generative AI whilst developing their academic skills.  

The University considers unauthorised use of generative AI to be a form of contract cheating. You can view the information students are given about contract cheating on our current students web area and there is support and guidance for students with the Study Skills AI learning resource. Our procedure for how to deal with suspected contract cheating provides information and guidance on steps to take if you are concerned about inappropriate use of AI tools in assessment. You can also speak to your School or Faculty Education Director (SED/FED) or your Academic Integrity Officer, as per your School’s Academic Integrity process. It is important that cases of suspected misuse are highlighted so that, as an institution, we can improve our understanding and knowledge of the area.  

Do not use Turnitin AI writing detection as your primary or sole indicator in a suspected case of cheating. You may want to use it to support you in decision-making once you have already determined that a breach in academic integrity has taken place but be extremely cautious about the number values and the risk of false positives. The use of Turnitin AI writing detection is under review this year by key stakeholders in faculties and schools, in partnership with the DEO.  

You should not upload student work to any external generative AI tools in the process of investigating suspected misconduct because of data protection. 

Study Skills’ Using AI at University tutorial provides a starting point to learn the basics about AI.  

Sign up to the regular BILT briefing to get notifications of the latest news, events, resources and training on generative AI. For informal and general queries, or to find out what others are using AI for, staff are encouraged to join the Artificial Intelligence for Teaching Learning and Assessment Teams Group.  

If you need discussion or support beyond the resources hosted by BILT, Study Skills, AQPO and DEO, we will endeavour to provide that for you. This includes supporting curriculum review, collaboration advice, skills development support for students, and support for curriculum development or innovative AI practice. Please email ai-education@bristol.ac.uk including your unit’s code and a brief description of your question, and the support you require.  

The University’s Isambard-AI is the most sustainable supercomputer in the UK and will help research the climate-emergency. Most accessible generative AI tools, however, consume very high levels of water and electricity. Some estimates claim that each generative AI search uses five times the energy of a conventional web search, and that each search consumes the equivalent of a bottle of water. 

With the rapid growth of the technology, AI data centres are developing at speed and are not presently designed with environmentalism in mind. There are many solutions available such as closed water-cooling systems and use of renewable energy sources and it is hopeful that these will be employed in the future. Some countries are redressing these problems through legal mandates. 

The demand for increased AI-capacity also contributes to unsustainable mining practices in the search for critical minerals and rare earth elements. These mining practices can also include illegal mining and labour exploitation and can contribute to socio-political tensions and conflict. As the technology advances it also creates electronic waste that often contains mercury and lead. 

Due to the current negative environmental impacts caused by many generative AI tools, you may elect to avoid or minimise use of AI.