Policy paper

Ofsted's approach to artificial intelligence (AI)

Published 24 April 2024

Applies to England

Background

AI-enabled tools and products are now more widely available in schools and other providers. This means that these technologies are affecting children and learners and the people who care for, educate and support them.

The Department of Education’s recent call for evidence about AI in education  demonstrates that many education providers are using AI to augment how they work.

For many education and social care providers this new technology presents both new opportunities and risks.^{[footnote 1]} The use of these tools will in many cases affect children and learners.

Ofsted is also using AI, most notably in our risk assessment of good schools to help decide whether a school that was judged good at its last inspection will receive a graded or ungraded inspection. We are also exploring how AI can help us to make better decisions based on the information we hold, and to work more efficiently.

This document outlines how Ofsted will use AI responsibly and fairly. It also sets out our position on the use of AI by education and social care providers.

Notes

Throughout this document we use ‘artificial intelligence’ as an umbrella term for a range of algorithm-based technologies that solve complex tasks by carrying out functions that previously required human thinking alone.

Ofsted’s position on the use of AI is consistent with the National AI Strategy. This aims to make sure that AI benefits all sectors and regions and that the UK gets the national and international governance of AI technologies right.

How we will use AI

Ofsted uses AI to make the best use of the information we hold. We are also exploring the potential for AI to further improve the way we inspect and regulate.

AI will help us work according to the values set out in our strategy: to put children and learners first while remaining independent, accountable, transparent and evidence-led. 

The areas where AI could have the biggest benefits for our work include:

• assessing risk
• working more efficiently through automation
• making best use of our data (particularly text) to generate new insights.

We also recognise the emerging risks of AI. We will follow the 5 principles set out by the AI regulation white paper. Table 1 describes some of the ways that we will comply with these principles.

Table 1: What we will do in relation to our use of AI

Regulatory principle	We will…
Safety, security and robustness	Make sure AI solutions are secure and safe for users and that they protect users’ data Continually test our AI solutions to identify and rectify bias and error
Appropriate transparency and explainability	Be transparent about our use of AI, and make sure we test solutions sufficiently to understand the decisions it makes
Fairness	Only use AI solutions that are ethically appropriate – in particular, we will fully consider any bias relating to small groups and protected characteristics at the development stage and then monitor it closely and correct it where appropriate
Accountability and governance	Provide clear guidance and rules for developers and users of AI within Ofsted about their responsibilities
Contestability and redress	Make sure that staff are empowered to correct and overrule AI suggestions – decisions will be made by the user, not the technology Continue to manage concerns and complaints through our existing complaints procedure

We will also follow the Central Digital and Data Office and Office for Artificial Intelligence (OAI) guidance on using AI in the public sector. This guidance sets out how to use AI ethically, fairly and safely.

Critically, we will never allow AI to impede our ability to inspect fairly and impartially.

It is vital that AI does not undermine either our inspectors’ judgements or our ability to respond flexibly and empathetically to the concerns of the public or providers. 

Providers’ use of AI

Ofsted aims to improve lives and raise standards in education and children’s social care. We do this through inspection, regulation and our insights.

Our inspection frameworks set out how we inspect providers. They explain our inspection principles, the main judgements that inspectors make, and, along with regulations, explain the areas that we take into account when inspecting providers. Our registration guidance and enforcement policies set out how we regulate.

In this context, we will consider a provider’s use of AI by the effect it has on the criteria set out in our existing frameworks.

Our inspection frameworks and regulations enable us to assess AI’s impact on children and learners as part of our wider inspection and regulatory processes.

Importantly, Ofsted will not directly inspect the quality of AI tools. It is through their application that they affect areas of provision and outcomes such as safeguarding and the quality of education. Leaders, therefore, are responsible for ensuring that the use of AI does not have a detrimental effect on those outcomes, the quality of their provision or decisions they take.

Ofsted supports the use of AI by providers where it improves the care and education of children and learners. We recognise that these tools can help providers make better-informed decisions, reduce workload and lead to innovative ways of working.

Table 2 describes some of the ways we expect providers to comply with the principles in the AI regulation white paper.

Table 2: What we expect from providers in relation to their use of AI

Regulatory principle	Providers are expected to…
Safety, security and robustness	Assure themselves that AI solutions are secure and safe for users and protect users’ data Ensure they can identify and rectify bias or error
Appropriate transparency and explainability	Be transparent about their use of AI, and make sure they understand the suggestions it makes
Fairness	Only use AI solutions that are ethically appropriate – in particular, we expect providers to consider bias relating to small groups and protected characteristics before using AI, monitor bias closely and correct problems where appropriate
Accountability and governance	Ensure that providers and their staff have clear roles and responsibilities in relation to the monitoring, evaluation, maintenance and use of AI
Contestability and redress	Make sure that staff are empowered to correct and overrule AI suggestions – decisions should be made by the user of AI, not the technology. Allow and respond appropriately to concerns and complaints where AI may have caused error resulting in adverse consequences or unfair treatment.

Building our understanding

Given the pace of change, the effect these new technologies are having on the lives of children and learners is still poorly understood. Accordingly, we will seek to better understand the use of AI by providers and emerging research on its impact on the outcomes and experiences of children and learners. We will do this by keeping up to date with relevant published research and communicating with providers so that we can learn from their experience of AI in their settings.

By better understanding the effect of AI in these settings, we can consider providers’ decisions more effectively as part of our inspection and regulatory activity.

Correspondingly, we will develop inspectors’ knowledge about AI so that they have the knowledge and skills to consider AI and its different uses.

Working with other regulators

We recognise that we are part of a wider system of regulation. Ofsted will continue to work with other regulators and central government departments, including the Department for Education and the Department for Science, Innovation and Technology, to ensure that our approach remains consistent with the latest guidance.

Through the Institute of Regulation and regular communication with other regulators, we will also ensure that the approaches we take to regulating sectors that use AI complement each other.

1. Technology is referenced as a source of risk in ‘Keeping children safe in education: statutory guidance for schools and colleges’, Department for Education, September 2023 ↩