The Academic Integrity Policy and the Procedures for Managing Allegations of Academic Misconduct, available on the UL Policy hub, indicate that it is a breach of academic integrity if AI is used in an unauthorised manner.
In other words, it is the responsibility of the module leader/equivalent to decide and communicate the level of acceptable use of AI to students i.e. to determine what is authorised/unauthorised use. If AI is used in a manner which has not been permitted by the module leader, then this is a breach of academic integrity.
Educators have autonomy to innovate with regards their approaches to teaching, learning and assessment. We would encourage you to take into account the UL Generative Artificial Intelligence Principles for the University, Staff, Researchers and Students and . The level of acceptable AI use is likely to differ considerably between academic disciplines and perhaps within disciplines. This is also likely to change within the months and years to come, as AI becomes further embedded within the workplace and our day-to-day lives.
The National Forum for Teaching and Learning are collating a repository of information on uses of AI within Further and Higher Education in Ireland: . This is a live resource so it might be helpful to check-in regularly.
There are a number of options available to you in terms of assessment. These can be summarised as follows:
- No adjustments/revisions applied to assessment
- Declarative/Discursive approaches
- Structural changes
The declarative/discursive/structural framework is best described in a recent paper authored by Thomas Corbin and Philip Dawson ().
These approaches focus on requiring students to declare and/or describe the level of AI use within their work. This approach is widely used internationally and can be helpful for educators as they support clear communication to students as to what is acceptable in the context of assessment.
- Declarative approaches, require students to include an acknowledgement or declaration template as part of an assessment, to indicate the nature and extent of AI use (including declaration that AI was not used in any aspect of the assignment). The Academic Integrity Unit have developed such . Your Faculty may also have similar declaration statements.
- Discursive approaches require students to indicate, often through commentary and/or appendices, the level of AI use within an assessment. The AI Assessment Scale is an example of such a discursive approach that might be of assistance to educators   Please note that an  is available.  developed by the authors of the AI Assessment Scale, signposts to papers which describe the application of the scale in practice.
- Academic staff should consider whether assessment rubrics need to be reviewed to ensure that marks are not being awarded for work that the Gen AI tool has completed e.g. grammar and punctuation.
There are a number of possible limitations to the declarative/discursive approach:
- Difficult to enforce (but not impossible): There is an implicit trust and transparency that is associated with declarative/discursive approaches. Breaches of academic integrity can be detected but can require significant investigation.
- Differences may emerge between modules in terms of use of discursive approaches e.g. deciding to use the AI Scale. This can cause confusion amongst students and educators.
This is a review of the types of assessment that are used within a module/programme and a redesign of these assessments where necessary. Structural changes may be needed for many reasons including:
- Concerns about security, authenticity, validity, reliability, feasibility or acceptability of the assessment
- Requirement of professional and statutory regulators who (re)accredit a programme of study
- New evidence emerging on innovations in assessment practices
A programmatic approach to structural assessment redesign may ultimately be the most robust approach for thinking about assessment in a holistic and meaningful way for your programme of study. This resource from JISC outlines the many different approaches to assessment in an AI enabled world: .
In all these cases, it may be necessary for you to reflect on the key considerations relating to the assessment:
- validity
-reliability
- security
-feasibility
- authenticity ()
-accessibility (Universal Design for Learning | ¾ÅÉ«ÊÓƵ)
-impact of assessment (assessment for/of/as learning) ()
Since the widespread availability of GenAI and the apparent threat it poses to academic integrity, there has been an explosion in the number of papers, resources, podcasts etc on this subject. Unfortunately, some guidance is not always helpful and can indeed contradict that which is endorsed by bodies such as the National Academic Integrity Network and other trusted international agencies/bodies. We would therefore recommend that educators and students exercise great care when deciding to accept guidance or information from a specific source. If in doubt, please reach out to the Academic Integrity Unit to assist and advice.
The Academic Integrity Unit and the Centre for Transformative Learning, regularly provide workshops on assessment redesign throughout the academic year. Please keep an eye-out for these workshops through your University email and on the UL Connect site.
The following might be helpful sources of information:
It is exceptionally challenging to ‘detect’ or ‘spot’ GenAI use. Rather, we encourage you to consider the points we have highlighted around assessment redesign above.
There has been some limited success with identifying unauthorised use of GenAI by:
- Identifying false citations or references within an assignment, possibly as a result of AI hallucinations
- Lack of citation
- A notable difference in writing and/or critical reflection/analysis in a particular submission relative to other assessment/communications
- Instructions/prompts intended for the GenAI left in the text
Relying on features such as:
- Buzz words/phrases or grammatical features
- Flow or structure
These features may provide some evidence on which to base a claim of academic misconduct. However, further investigation is required in order to proceed with an academic misconduct investigation. It may not always be possible to proceed with a case based on the information available.
There have been many documented cases of academic misconduct decisions being overturned, following an unfounded allegation of unauthorised use of AI: .
The AI Advisory Council recommend that detectors should not be used () and NAIN  also recommend that they should not be used:
"Do not rely on GenAI ‘detection systems’. None of the tools which are currently available are fully capable of detecting the use of GenAI (except in the most obvious cases which may also have been identified by expert reading and scrutiny) and may also lead to ‘false positives’ (incorrectly concluding that human-written text was AI-generated) and difficult-to-interpret scoring. Detection systems cannot be relied upon to detect use of GenAI accurately or consistently. In addition, there may be serious data protection, privacy, and intellectual property concerns in the use of any such tool, particularly if it has not undergone appropriate approval by institution. Turnitin’s detection tool is available in some institutions, but users should be aware of concerns about its capabilities in terms of more recent versions of GenAI, a reported high rate of ‘false positives,’ and some ambiguity on how to interpret its results". 
We urge staff and students not to upload student assignments/assessments to any websites or tools to check for AI use.