Drawing the line: AI use in academic and professional assessment

The rapid advancement of Artificial Intelligence (AI) technology is driving new possibilities and aspirations across the education and assessment sector.

As generative AI technology and the regulations to govern its use continue to develop, how do we know how far we can go in the use of generative AI in marking?

What guidance is there to ensure integrity and fairness, security and accuracy for the use of AI within assessments, whilst providing rich and insightful feedback to support the academic or professional learner journey?

In this blog post, we explore how and where we draw the line in the use of AI in the academic or professional assessment process.

A framework for ethical integration

On 13th January 2025, the UK Government published an independent report, AI Opportunities Action Plan. Its aim was to set out how it would ramp up AI adoption across the UK to boost economic growth, provide jobs for the future and improve people's everyday lives. It is looking to proactively shape the AI revolution, rather than waiting and letting it shape Britain.

The Government said it has taken decisive action to support the AI sector and take down barriers to growth, introducing transformative planning reforms to make it easier to build data centres and for Skills England to help ensure people are prepared for jobs in the AI-powered industries of the future. The Dept of Science, Technology and Innovation aims to drive technological transformation to support AI to directly benefit working people by improving health care and education.

Specifically for education, the report said it anticipated that AI would be a useful tool for assessment in the education sector, giving the example of the Department for Education’s own Proof of Concept (PoC) with a generative AI and rules-based marking tool that showed 92% accuracy in a pilot with teachers on year 4 literacy work when drawing from appropriately coded educational data and content¹.

Despite the government action plan, the real challenge is that the line has not yet been drawn as to how, and exactly where AI can be used within academic and professional assessment ethically.

Guiding teachers and learners

The Artificial Intelligence Scale (AIAS): A Framework for Ethical Integration of Generative AI in Educational Assessment² by Perkins et al sets out a scale ranging from absolutely no AI use in assessment all the way through to full integration of AI. Its purpose was to empower educators to select the appropriate level of generative AI usage in assessments based on the learning outcomes they are looking to address.

The AIAS was also created to offer greater clarity and transparency for students and educators, provide a fair and equitable policy tool for institutions to work with and to give practical, quick and easy examples of how AI can be implemented to address the uncertainty and often negative connotations of AI’s use within education. Considering a scale is one way to look at where the line may be drawn.

Key education institutions have also started to give their teachers and learners guidance on the use of AI within assessment.

For example:

• London School of Hygiene and Tropical Medicine has issued a Statement on using Artificial Intelligence (AI) in Assessments³
• University of Oxford’s online Centre for Teaching and Learning has an AI and academic practice resource which includes An introduction to the use of generative AI tools in teaching: Ten ideas on the use of AI to support students in their learning, together with suggested responses for teaching⁴

In their guidance, both institutions have highlighted the potential positive influence of AI, suggesting ways in which it can be used to support learners, teachers and assessment practices.

The EU AI Act: Pioneering AI Regulation

The EU AI Act⁵, the world's first comprehensive AI regulation, was published on July 12, 2024, and came into force on August 1, 2024. It will be fully applicable by August 2, 2026, with certain prohibitions and AI literacy requirements already in effect since February 2, 2025. This Act overlaps with the GDPR in areas related to personal data.

At RM’s Spring Conference, Dr. Robert Harrison, an attorney, highlighted the Act's goal to create a human-centric, trustworthy, and risk-based AI system. This regulation aims to balance innovation with ethical responsibility, addressing the potential misuse and unintended harm of AI technologies in daily life and business.

Key Points from the EU AI Act:

• Human Rights Protection: Recital 48 emphasises the protection of fundamental rights, including the right to education, from high-risk AI systems.

• Educational Impact: Recital 56 underscores the importance of AI in promoting high-quality digital education and training, while also classifying educational AI systems as high risk due to their potential impact on a person's life and livelihood.

Risk Classification Framework:

The Act categorises AI systems into four risk levels: unacceptable, high, limited, and minimal, each with specific regulatory requirements. High-risk AI systems, which affect safety or fundamental rights, are divided into two categories:

1. AI systems in products under EU product safety legislation (e.g., toys, aviation, cars, medical devices).

2. AI systems in specific areas requiring registration in an EU database (e.g., education, law enforcement, employment).

All high-risk AI systems must be assessed before market entry and throughout their lifecycle.

Supporting Innovation:

The Act supports AI innovation and start-ups in Europe by allowing companies to develop and test general-purpose AI models before public release⁶. National authorities are required to provide testing environments that simulate real-world conditions, helping SMEs compete in the EU AI market.

Dr Tom Chatfield, Author and Tech Philosopher, referring to Robert Harrison’s presentation, suggested the assessment community should agree a code of practice by mutual consent and stick to it. “That forms the basis for ongoing conversations about where these boundaries should lie, what innovation looks like,” Tom explained. “I think that's important, because otherwise you get into these various arms races and then you shut down innovation because then crucially, mutually you can agree to adapt it, to expand it, to extend it and to enforce it when needed as well.”

The line is still being drawn

It is apparent that although artificial intelligence is 70 years old this year, there is still a long way to go in defining how and where to draw the line in the use of generative AI in the academic or professional assessment process.

“Something that worries me tremendously is that if we're not very careful about it, if we don't bring elected politicians along with us, that policies will be imposed upon the assessment community, upon the educational community, with which we will have difficulty living with,” said Robert Harrison.

Working together as a community to explore how AI could be used ethically within assessment to improve learning outcomes is key.

As the codes of practice, the law and the technology continue to develop, RM is taking a long-term, proactive view, exploring how AI could be used ethically in assessment to drive:

• better learner outcomes without diminishing the vital role of human expertise

• continuous improvement without introducing risk

• support and collaboration with the wider community to ensure the integrity, fairness, security and accuracy of assessment marking

We are collaborating on a range of Proof of Concepts (PoCs) with our customers and prospects, providing academic and professional assessments, who will gain early access to innovative AI-driven solutions and explore how AI marking can be used safely and securely to reduce workload, enhance accuracy, and deliver personalised feedback to enrich the lives of learners.

RM aims to give confidence in the use of AI within assessments to exam boards, awarding bodies, politicians, learners, and parents everywhere.

To continue this collaborative journey, we are delighted to announce our next Bridging AI in Assessment conference, where we'll start to put AI powered assessment into action.

This event will take place in London, on Wednesday 8th October.

We'll be sharing full details in due course, but for now, please register your interest so we can keep you updated.

Sources

¹Department for Education, ‘Use Cases for Generative AI in Education - Building a proof of concept for Generative AI feedback and resource generation in education contexts: Technical report’, 2024 (accessed 03 December 2024)

²The Artificial Intelligence Assessment Scale (AIAS): A Framework for Ethical Integration of Generative AI in Educational Assessment. (2024). Journal of University Teaching and Learning Practice, 21(06) https://doi.org/10.53761/q3azde36

³LSHTM Statement on using Artificial Intelligence (AI) in Assessments Version 1.1 (Updated September 2024) https://www.lshtm.ac.uk/media/74671#:~:text=If%20AI%20tools%20are%20used,AI%20tools%20is%20given%20below

⁴An introduction to the use of generative AI tools in teaching: Ten ideas on the use of AI to support students in their learning, together with suggested responses for teaching, Centre for Teaching and Learning, University of Oxford https://www.ctl.ox.ac.uk/ai-tools-in-teaching

⁵Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act) (Text with EEA relevance), https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R1689

⁶EU AI Act: first regulation on artificial intelligence, European Parliament, Published: 08-06-2023Last updated: 19-02-2025 - 17:46 https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence#more-on-the-eus-digital-measures-9