(1) Our approach to Artificial Intelligence (AI)

Warwick Law School is committed to developing a holistic approach to AI in all aspects of our education. We seek to develop AI literacy among our students to instil good practice regarding the responsible use of AI as a tool to assist (but not replace) their learning and skills development. This includes promoting awareness about the ethical use of AI as well as about the concerns over the robustness and bias of many AI systems. We will monitor the use of AI by students to ensure that this is done responsibly and ethically. This includes monitoring the impact the use of AI has on development of essential academic skills by our students. Academic integrity must be ensured in all activities involving the use of AI.

We are committed to understanding the implications of AI for our own subjects as well as the ethical issues raised in all aspects of AI and to ensurin that what we teach includes consideration of the legal and ethical challenges brought about by the use of AI.

(2) Artificial Intelligence (AI)

(a) AI Technologies

Artificial Intelligence is an umbrella term covering a wide range of algorithmic and data-driven software systems. There are different types of algorithms (and combinations of different approaches) with varying capabilities, either of a deterministic or adaptive machine-learning type. AI performs a wide variety of tasks, not always clearly identified as “AI” in people’s minds (e.g., spell-checkers, auto-complete, recommendations on shopping websites or streaming services, etc).

The Organisation for Economic Co-operation and Development (OECD) defines an AI system as “a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments. Different AI systems vary in their levels of autonomy and adaptiveness after deployment”.[1] All AI systems have in common that they offer varying degrees of automation for particular tasks, often more efficiently and quicker than humans.

A particular sub-category of AI is Generative AI Systems (“GenAI”). They can be useful for some academic tasks, but can also be misused in a way that violates our standards and expectations as to academic integrity.

GenAI, such as ChatGPT or Claude, is based on deep-learning neutral networks utilising “transformer” technology. They are trained on vast amounts of data and produces outputs based on that data. The robustness and reliability of outputs can vary dramatically, and there are many caveats to the use of generative AI systems.

(b) Limitations of GenAI

Importantly, AI does not process information in the way humans do. It does not develop knowledge. It lacks common sense and the ability to locate information in a wider context/experience. Its outputs are generally based on the data on which it was trained, and, in the case of “self-learning” (adaptive) algorithms, data acquired during deployment; it therefore lacks the ability to produce truly original and creative outputs. It does not have the ability for critical thinking – one of the key academic skills we seek to instil and develop in our students.

AI can be a very useful tool to help with certain tasks. It can uncover connections in data that might not be apparent to humans (although it is at risk of overfitting, or “hallucinating”, such connections) and can structure data better and faster than humans.

Recent advances in the performance and capability of certain AI systems, particularly generative AI systems, have renewed interest in this technology, and have led to an explosion in academic, policy and legislative work. A lot of unsubstantiated or exaggerated claims are made about AI, and it is sometimes difficult to get a clear sense of what the technology is capable of and what its limitations are. There are also concerns about the quality of the data on which AI systems are trained, particularly regarding its robustness and accuracy, and biases. Furthermore, the degree of energy use and the resulting environmental impact of the infrastructure (e.g., data centres) required for some AI systems is starting to attract more attention and concern.

(3) Permitted and prohibited uses of GenAI and other AI tools in the classroom

(4) Academic Integrity

With regard to academic integrity, our policy builds on the University’s Institutional Approach to Artificial Intelligence and Academic Integrity. Any changes to the University’s Institutional Approach, or any new University policies on the use of AI take precedence over this policy whenever there is a conflict between them.

In accordance with both the University’s and the Law School’s academic integrity policy, the Law School prohibits copying/paraphrasing either whole outputs or elements of outputs generated by a GenAI system and submitted by students as their own work. Similarly, the use of AI to complete certain parts of an assessment, such as analysis or evaluation, is not permitted.

However, even where the use of GenAI systems is not prohibited, the Law School discourages the use of such outputs even with correct and complete attribution. Assessments should be used by students to demonstrate their knowledge of the subject (i.e., what has been taught, including set readings) and their understanding and ability to utilise this in responding to specific questions. The use of GenAI in this process might make it more difficult for markers to establish how well a student has demonstrated this. The way in which such outputs are used as part of an assessment will be considered in grading the assessment against our marking criteria.

The use of an AI system for any aspect of completing an assessment must be disclosed by a student. A failure to do so constitutes an academic misconduct. Such a disclosure must cover the following points:

• Why was a GenAI system used (help in understanding the question; help in structuring work based on arguments developed; assist with initial summary of readings etc)?
• Which GenAI system or systems was/were used (e.g., ChatGPT, Claude etc)?
• How has the AI output been used in preparing the assessment? (as a research tool, to test arguments etc; see table below for acceptable uses).

The way AI generated outputs are used by you in drafting your assessments will be taken into account when grading the assessment against the generic grade descriptors and assessment-specific marking criteria.

(5) Permitted and prohibited uses of GenAI – Guidance to Students

Unless expressly required for a specific assessment task, Warwick Law School does not encourage the use of GenAI tools for assessments. Students should be aware that GenAI tools can weaken the quality of your work – particularly if you rely on GenAI for accuracy, relevance and rigour. You should be particularly mindful of the tendency of GenAI tools to cite non-existent resources or information (“hallucinations”). GenAI may generate inaccurate or otherwise poorly constructed arguments. GenAI does not have the ability to demonstrate critical thinking, nor the ability to be genuinely creative. Most importantly, assessments are there for you to demonstrate to us (and to yourself!) how well you have understood what you have studied.

Nevertheless, WLS acknowledges that some students may use GenAI technology when preparing for assessments and writing essays, and may want to gain experience of utilising it for their future legal careers. Below, we provide a clear rule and interpretative guidance on what would constitute tolerated and prohibited use of GenAI. This is always subject to specific instructions given for each assessment.

Rule and Guidance

The key overarching rule is: GenAI can act as a personal assistant for you (e.g. assistance in understanding key issues, help with research, proof-reading your work etc.). However, it must never be used to create the work, or certain parts of the work, for you. You must not submit any AI-generated output as your own work.

The table below summarises what is acceptable and what is prohibited in a bit more detail. Note that any use of GenAI may be prohibited for specific assessment tasks. This might be the case e.g., where the assessment tests your key legal or academic skills.Here, your own unaided work would be essential. Always check the assessment instructions before you use any GenAI tool!