Artificial Intelligence (AI) in schools: guidelines and guardrails

Section 5 - Key Definitions

To navigate the evolving landscape of AI, it’s important to become familiar with the terminology that surrounds these technologies. This section clarifies and explains essential terms, aiming to demystify AI and illustrate its relevance to modern education.

Adaptive Learning

Adaptive learning refers to educational methods and technologies that personalise learning experiences in real time, responding to each pupil’s strengths, weaknesses, pace, and preferences. Through the use of AI and data-driven insights, adaptive learning systems, overseen by teachers exercising professional judgement, continuously analyse pupils’ interactions and outcomes, adjusting the content, support, and level of challenge to suit individual needs.

This approach aims to maximise engagement, promote mastery, and ensure that all children and young people have equitable opportunities to succeed, no matter their starting point or approach to learning.

Artificial Intelligence

The Scottish AI Alliance offer two possible definitions of Artificial Intelligence, while recognising the challenge of defining the term, as there is not a single agreed definition.

The first definition put forward describes artificial intelligence in the following way:

“Artificial Intelligence is the field of study within computer science that focuses on designing computer systems that can perform tasks that we would normally consider to need ‘human intelligence’, such as making decisions and producing natural language.

“The technologies that result from this field of study are known as ‘AI technologies’.”

While the second definition, which is based on the OECD definition, states:

“An AI system is a machine-based system that can for a given set of human-defined explicit or implicit objectives, draw conclusions, from the input it receives, how to generate results such as make predictions, content, recommendations, or decisions that can influence physical, real, or virtual environments.

“Different AI systems are designed to operate with varying levels of autonomy and adaptiveness after deployment.”

Both definitions have strengths and weaknesses, which the Scottish AI Alliance explore in detail.

Generative AI

In recent years we have seen the growth of Generative AI systems, often referred to as ‘GenAI’. Generative AI is an AI system that generates text, images, audio, video, or other media in response to user prompts. This includes systems like ChatGPT, Google Gemini or Microsoft Copilot.

Generative AI uses machine learning techniques to create new data that shares characteristics with its training data, often producing outputs that are nearly indistinguishable from human-created content, which can present ethical challenges.

Generative AI responds to user prompts to generate tailored outputs like text, images, or videos. It does this by using large amounts of data to find patterns and connections, with the help of advanced computer programs called neural networks.

When a prompt is given, the AI predicts and assembles the most likely sequence of words, pixels, or sounds based on what it has learned from its training data. This process enables GenAI to produce content that often closely mirrors human-created material, continually developing as it processes more data and receives feedback.

As GenAI is essentially predicting the ask of the user based on previous data that it has been trained on, it can be significantly subject to bias. This is where the importance of ethics in relation to AI must be emphasised.

This resource by Education Scotland explains some of these key terms like ‘Generative AI’ and ‘Machine Learning’ in more detail.

Ethical AI

Ethical AI refers to the practice of designing and deploying artificial intelligence systems in a way that is fair, transparent, accountable, and respectful of human dignity, values and rights. It involves establishing principles and techniques to prevent potential harms like bias, discrimination, and invasion of privacy, and to ensure that AI technologies benefit society while mitigating risks.

The Scottish AI Alliance defines Ethical AI as “aligned to the United Nations Sustainable Development Goals and Scotland’s National Performance Framework, respecting and supporting our human rights, our environment and our communities.

“Ethical AI is accountable to the people who are affected by it and respects the laws and rights of the people of Scotland as well as internationally.”

In an educational context, this means ensuring that AI tools and systems used within schools promote fairness, transparency, and inclusivity, and that they safeguard the rights, privacy, and well-being of all children and young people and teachers.

Further information on the key considerations underpinning Ethical AI can be found in Section 5.

AI Literacy

Learning about and learning with AI requires that teachers, children and young people develop their AI literacy skills.

The OECD defines AI literacy as: “the technical knowledge, durable skills, and future ready attitudes required to thrive in a world influenced by AI. AI literacy enables children and young people to engage, create with, manage, and design AI, while critically evaluating its benefits, risks, and ethical implications.”

While this guidance document uses the term ‘AI literacy’ to describe the emerging knowledge, skills and attitudes in relation to AI, it should be noted that children and young people have expressed reservations about using the term ‘literacy’ for other topics, arguing that it is confusing. Therefore, the term AI literacy should be avoided in a classroom context and replaced with ‘Learning with and about AI’.

With the growth of AI across our society, being able to use and understand AI is fast becoming an essential facet of knowledge and skill. To ensure that children and young people can contribute to society, participate in the economy where AI and the need for digital skills are becoming increasingly prevalent and have their voices heard in societal debates about AI, there must be an attempt to appropriately integrate AI literacy across the curriculum.