Why AI training should be mandatory in every school and college

Most schools would not dream of saying that safeguarding training is optional.

We do not train only the safeguarding lead. We do not offer training solely to those who are particularly interested. We do not assume that staff who attended a course three years ago still know everything they need to know. Safeguarding is everyone's responsibility.

AI should increasingly be viewed in the same way.

That does not mean artificial intelligence presents the same risks as safeguarding. It means the principle of shared responsibility is the right one. AI is already being used by teachers, support staff, leaders and pupils. It is influencing how work is produced, how information is found, how decisions are made and how learning takes place.

Every member of staff therefore needs a basic level of AI literacy.

The age of the AI enthusiast is over

For the past few years, AI in schools has often been led by enthusiasts. A digitally confident teacher experiments with a chat bot. A senior leader attends a conference. A member of staff shares some useful prompts. An AI working group is established. This has been an important stage in the sector's development. But it is not a sustainable model for what comes next.

AI is no longer a specialist interest. It is becoming part of the everyday educational environment.

The question is no longer whether a member of staff chooses to use AI personally. They are working in a world in which pupils are using it, colleagues are using it and AI is increasingly embedded within the platforms, products and services used by schools.

We would not accept a member of staff saying, "I don't really do safeguarding." Increasingly, "I don't really do AI" cannot be an adequate professional position either.

Training must be about more than prompts

Much early AI training has focused on prompting. That is understandable. Knowing how to communicate effectively with generative AI can be useful. But prompting is not AI literacy.

Staff need to understand why AI can produce convincing but inaccurate information. They need to know what happens to the information they enter into an AI tool. They need to understand bias, intellectual property, data protection and the importance of human oversight.

They need to know when AI can be useful - and when it should not be used. Most importantly, they need the confidence to apply that understanding to their own work. The purpose of good AI training should not be to create people who can remember twenty clever prompts. It should be to create professionals who can make informed judgements.

The risk of a two-tier workforce

One of the greatest risks facing schools and colleges is not that nobody uses AI. It is that a small group of confident early adopters races ahead while everyone else is left behind.

We are already seeing this divide emerge. In the same organisation, some staff are using AI regularly and transforming aspects of their work, while others have had little or no training and may not understand even the basic opportunities, limitations and risks.

That is neither equitable nor safe. It can lead to inconsistent practice, different approaches to sensitive data, uneven expectations for pupils and significant variation in how AI is used across classrooms and departments. It also means that access to the benefits of AI may depend on which teacher, subject or school a young person happens to encounter.

Mandatory training creates a common floor of understanding. It does not mean that every member of staff must use the same tools or become an AI expert. Nor does it mean removing professional judgement. It means ensuring that everyone starts with the knowledge they need to make informed decisions.

Schools and colleges should then build on that common foundation. Staff need opportunities to apply their learning to their own roles, share effective practice and develop their confidence over time. Some will become champions and innovators. Others will use AI more selectively. But nobody should be excluded from the basic knowledge required to work safely and confidently in an AI-shaped education system. The aim is not uniformity. It is shared understanding. That is why AI training can no longer be left to the enthusiasts.

A common foundation for every member of staff

This is why we believe every school and college should establish a minimum entitlement to AI training.

Everyone should understand:

• what AI is and what it is not;
• what generative AI can and cannot do;
• how to interact with AI effectively;
• how to protect personal and sensitive data;
• why AI outputs must be checked and verified;
• how bias and misinformation can arise;
• where human judgement remains essential; and
• the organisation's expectations for safe and appropriate use.

This should apply to teachers, leaders and support staff.

The risks do not disappear because someone does not teach in a classroom. AI may be used to draft communications, summarise documents, prepare resources, support administration or analyse information. Safe and responsible use is an organisation-wide responsibility.

From one-off training to annual refreshers

There is another lesson we can take from safeguarding: training cannot be completed once and forgotten. AI is developing too quickly. The tools are changing. Their capabilities are changing. Regulation and guidance are changing. The ways pupils use AI are changing. The risks are changing.

A training session delivered two years ago cannot prepare staff adequately for the AI environment they are working in today. We therefore believe schools and colleges should move towards annual AI training for all staff, supported by regular updates during the year.

The annual refresher should revisit the fundamentals, reflect changes in technology and guidance, and use real examples from the organisation's own experience. It should ask not simply, "What do you know?" but, "What has changed in your practice?"

Training needs a framework

Training should sit within a wider strategy. A school can train every member of staff and still have an incoherent approach to AI. Without clear policies, leadership, appropriate tools, curriculum thinking and a shared understanding of effective pedagogy, training risks becoming another isolated initiative.

This is why it is important to consider AI implementation as a whole-organisation endeavour with interconnected pillars that schools should develop together: AI Literacy; Policies and Ethics; Tools and Systems; Digital Pedagogy; and Collaboration and Community. These pillars provide a useful structure for thinking about sustainable and effective AI adoption, recognising that progress in one area alone is unlikely to deliver meaningful change.

Training is not the whole answer, but it is the foundation on which much of the answer depends.

Choosing the right professional development

Schools now have access to a growing range of AI professional development opportunities, and the quality and focus of these programmes varies considerably. The most effective training is practical, education-focused and helps staff understand not only what AI is, but how it can be used safely, responsibly and effectively within teaching, learning and school operations.

Training should follow a structured learning pathway that takes educators from the fundamentals of AI through to its practical application in education. It should also help build a shared understanding across staff, ensuring that AI is implemented consistently rather than relying on isolated pockets of expertise.

Ultimately, whichever training a school chooses, it should contribute to wider organisational development. Professional learning is most effective when it supports a coherent whole-school strategy that aligns with these wider priorities, ensuring training is not an isolated activity but part of sustained, joined-up change.

Everyone's responsibility

We are reaching the point where AI literacy should be considered part of the basic professional competence of working in education. Not everyone needs to become an AI expert. Not every teacher needs to use AI in every lesson. Not every school needs to chase every new tool. But everyone needs to understand enough to act safely, make informed decisions and support young people in an AI-shaped world.

The greatest risk is not that schools fail to train a handful of AI experts. It is that they create a growing divide between confident early adopters and everyone else. A school in which some staff are transforming their practice with AI while others do not understand the basic risks is neither equitable nor safe. Mandatory training creates a common floor of understanding.

AI training should be mandatory.

It should be practical.

It should change behaviour.

And, like safeguarding, it should be everyone's responsibility.

‍

‍