Generative Artificial Intelligence - information for centres
We must ensure equity and fairness for all learners studying our qualifications.
For session 2023–24, learners are not permitted to use generative AI tools to create outputs – for example text, prose, formulae, code, images, video, audio – that they then submit as their own work for assessment tasks that contribute towards an SQA qualification. These tasks include exams, unit assessments, coursework, and portfolios. Doing so would constitute plagiarism and could result in awards being cancelled.
To support centres in ensuring learners’ work is their own, we have produced a guide to authenticating learners’ work.
Additionally, learners must not include outputs from generative AI tools that are referenced as a source for assessment tasks that contribute towards an SQA qualification. There are currently some significant issues regarding the reliability and validity of these outputs that mean referencing the tools could be inappropriate or disadvantageous to learners.
We understand the need to embrace the opportunities that new technology offers and will continue to keep our position under review. Any changes to assessment practice will be based on evidence (including the views of centre staff and learners).
The following sections provide further information on the reasoning behind our current position on the use of generative AI tools.
There is evidence, acknowledged by the creators of generative AI tools, that outputs from these tools can be biased, incorrect, and can fabricate information. Outputs can also be inconsistent, even when using the same prompts – making it difficult for assessors to authenticate the sources. For these reasons, outputs from generative AI tools are not currently considered valid or reliable. Learners studying towards SQA qualifications should use valid, reliable and authoritative sources of reference to support their work.
Using outputs from generative AI tools as sources may not meet the referencing requirements of specific courses and could impact the number of marks a learner can achieve. For example, some SQA qualifications require sources to be recent and text output from generative AI tools can be difficult, or impossible, to date.
An important factor which could impact equity and fairness for learners is the age restriction that the creators of generative AI tools have placed on their products.
Looking to the future
We understand that as technology advances, the barriers to using technology and the current flaws within a technology can be overcome. We acknowledge that in the future, generative AI will have a much greater impact on all our lives, including education and assessment. SQA welcomes the recommendations for AI use in the recently published Independent Review of Qualifications and Assessment and we look forward to working with others in the education sector. Together we must explore opportunities while balancing the need to mitigate risks to the integrity of SQA qualifications and assessments, as well as ensuring equity and fairness for all learners.
SQA will continue our work on the use of generative AI tools in the assessment context and look forward to their continued evolution. We will be engaging with centre staff and learners around AI and will publish further guidance where appropriate.
Thank you for your ongoing co-operation.