Generative AI for students

Introduction

It seems that everyone is talking about Generative artificial intelligence (Generative AI) and how it will change the world (see below for an explanation of what we mean by Generative AI). As a university we not only recognise that Generative AI offers incredible potential to develop new ways of learning, but also that it can generate false information, discriminate against some people and could be misused to cheat. Explaining how Generative AI will affect your study is challenging because the technology is advancing so quickly; as it continues to evolve, so will our thinking. This document outlines the OU’s current position on acceptable and unacceptable uses of this technology.

What is the Open University's experience with artificial intelligence?

Since 2012, the Open University has a strong track record of researching and using artificial intelligence (AI) to enhance your student experience (see, for example, OU Analyse). AI is embedded in much of our everyday technology, such as accessibility software, some search engines, and some chatbots. We are now exploring and embracing the opportunities Generative AI presents for enhancing learning, teaching, and assessment at the OU, and will navigate the challenges and risks thoughtfully. Our latest research, scholarship, and innovation in this area will inform how we use Generative AI to enhance your study experience. Importantly, we will continue consulting with students, listening to your ideas and concerns to ensure we are inclusive, innovative, and responsive.

What do we mean by Generative Artificial Intelligence?

Recent excitement about computers producing original text, pictures, movies, music and voices has been driven by the development of easy-to-use, publicly available ‘Generative AI’ software. Examples include OpenAI’s ChatGPT and DALL.E2, Stable Diffusion, Google’s Bard and Generative AI-assisted search engines. You can provide these tools with a prompt about almost any topic and they will respond in fluent English as well as many other languages, often sounding surprisingly knowledgeable, or with an image or with computing code. Although outputs produced by Generative AI tools look plausible, they are often misleading, made up, or may be entirely wrong.

So how does it work? Generative AI is a category of AI that involves training machine learning models on extremely large datasets and has actually been around since the 1960s. Generative AI tools know which words and phrases are most frequently found with one another, but they do not know what any words or phrases mean. So, at a very basic level, when producing text, the Generative AI program is simply using statistics to choose the most appropriate next word or phrase. It’s like predictive text, except the algorithms are more sophisticated.

Using Generative AI in your studies

In this section we will consider when it’s okay to use Generative AI and associated tools in your studies and what is inappropriate use.

When used appropriately, Generative AI and associated tools can be used to support your learning and enable you to develop the skills you’ll need in an evolving digital world. However, Generative AI software can make mistakes, sometimes critical ones, so it’s important to use it in combination with a variety of reliable sources, such as your module materials and library resources. Overreliance on Generative AI tools will reduce your opportunities to develop essential skills in writing, critical thinking, and evaluation skills that you will need throughout your studies and in everyday life, including in the workplace.

Here are some ways you could use Generative AI in your studies:

  • To enhance accessibility and inclusion, for example, through captioning and audio description, text-to-speech and speech-to-text and assistive technologies.
  • To help improve writing skills, for example by correcting spelling/grammar mistakes. Tools such as Microsoft Word grammar checker and Grammarly already do this and can be more appropriate for assessed work.
  • To revise your own knowledge on a topic. This could be by asking it to summarise key information on a topic you have already studied (remembering to cross-reference the output with your module materials) or asking it to summarise a longer piece of text (excluding OU material*) to check your interpretation.
  • To support critical thinking skills; for example, you could ask it to challenge your key points to help you identify potential gaps and strengthen your arguments.
  • To develop understanding of complex concepts or apply discipline knowledge in new contexts.
  • To generate initial ideas; for example, if you’re struggling with writer’s block, it can be used to draft ideas, plans, or structures that you can adapt based on your own understanding of the topic.

Remember that you will need to critically evaluate any information generated by such tools, as Generative AI has no understanding of the accuracy of the information it generates and so can provide incorrect or false information.

The University requires that all work you submit is your own, with all sources fully attributed.  Therefore, if you use Generative AI in ways specified in the list above for your assignment, make sure you reference and acknowledge this according to the reference guidance on the Library website. For postgraduate research students, please refer to your supervisors.

It is best practice to include a copy of the Generative AI output as an appendix to your submission. For example, include:

  • Prompt author
  • Date
  • A description of the use of AI: “I used AI assisted Bing to generate initial ideas. I then asked it to dispute my key points once I had formulated them”.  
  • The following prompts were put into Bing: <list prompts>
  • The following output was generated: <include AI generated output>
  • I took the following actions: <explain how you adapted the output based on your own understanding of the topic>

Please check whether the assignment brief or your tutor require this.

*In accordance with the student code of conduct, you must not put any OU materials, including assessment questions, into Generative AI tools. Doing so may constitute a breach of the Open University copyright.

The University regulations on Academic Misconduct have not changed. We have a page describing our approach to academic conduct as well as some self-study resources on OpenLearn to help with common concerns about plagiarism and how to avoid it. Good academic conduct means you complete your academic work independently and honestly, using your own words and appropriate academic style, with all sources fully attributed according to academic requirements.  

You should be aware that inappropriate use of Generative AI (such as copying and pasting AI generated answers without referencing them) is prohibited by section 3.5 of our Academic Conduct Policy for all undergraduate and taught postgraduate students; and by sections 3 and 4 of the PGR Plagiarism and Research Misconduct Policy for postgraduate research students. (Find guidance on how to reference under the resource section below.)

Module teams and tutors/supervisors are vigilant about assessing whether students’ submitted work is written independently and with honesty. The University uses software to help identify potential cases of plagiarism, and tutors/supervisors and markers can spot changes in writing style that indicate your assignment may not be all your own work.

Where it is suspected that a submitted assignment has content that is passed off as a student’s own when it is not, cases are investigated. If the case is determined to be academic misconduct, you may receive a warning, a penalty or loss of marks. Any suspected cases of Generative AI plagiarism may result in a viva exam in which you will undergo a detailed interview about your knowledge of course material. In severe or repeated cases, you may be referred to the Central Disciplinary Committee which could result in penalties including total exclusion from the University. For postgraduate research students, please refer to the PGR Plagiarism and Research Misconduct Policy.  

It is your responsibility to ensure you behave with academic integrity and do not use Generative AI in an inappropriate manner. 

Here are some resources to support good academic conduct:

What’s next?

As outlined above, this guidance has been created to provide immediate support and direction for our students, but we know there’s more work to do. The OU is looking at ways we can embed Generative AI in our teaching to make sure that we equip you as students with the skills you will need for the future - but this is going to take time.

Although the specifics of how Generative AI is used in learning, teaching and assessment will be different according to the needs of each subject (for example, using Generative AI in a maths module might look very different to how it’s used on a creative writing course or a dissertation), we will be working with Faculties and Schools to develop a consistent approach as far as possible. We will also be working closely with the OU Students Association (particularly their Senate Reference Group and Student Leadership Team) to make sure that student voice is embedded in this work and will be looking for other ways to engage with you.

For now, if you have any questions on using Generative AI in your studies that haven’t been covered in this guidance, please contact your Student Support Team; or for postgraduate research students the Graduate School.