Implementing AI effectively to safeguard students

Artificial intelligence in education

Given its prevalence in daily life, AI is already being used in education to some extent. For example, students are using AI tools such as Chat GPT to support their learning, and schools and colleges may be using AI to detect plagiarism. We are also aware that in some schools and college teachers are using AI as a tool to generate teaching materials or to structure timetables.

The Department for Education expects schools and colleges to make the most of opportunities made available by AI, some of which include the ability to easily personalise learning by adjusting the format of a lesson or using predictive typing to make education more accessible, or AI laboratories supporting students where they are unable to participate in lab work on the ground. AI can also support teachers to work more efficiently, supporting with marking exams and automating administrative tasks. With the increase in SEND challenges in schools and colleges, it may also be used as a useful tool to adapt plans for SEND students.

Safeguarding risks of AI

Whilst there are many benefits to be had from integrating AI into schools and colleges, it does come with risks which must be considered to ensure appropriate safeguarding of students.

One of the risks of AI which many will be aware of is the creation of ‘deep-fakes’. These are AI created synthetic media which take an individual’s likeness, such as their face or voice, and use it to create realistic, but ultimately fake, content.

For example, AI can be used to generate a video of a person who does exist, using their face and voice, to create the impression that the person has said or done something which they haven’t. Alternatively, AI can be used to create wholly fake personas which appear to be real people but are completely AI generated. This is particularly prevalent on social media.

Whichever form of deep-fake is created, it is often used with malicious intent. The purpose could be to create inappropriate photos of people, to mimic a person’s voice on the phone to request a transfer of money for an ‘emergency’, or to blackmail people.

There are cases of students sadly taken their own lives after viewing negative online content as algorithms used by social media resulted in binge periods of content which can become very dark and graphic.

It should also be kept in mind that using AI without appropriate review processes can risk bias and discrimination. The data used to train AI systems is vast and, if that data is biased, AI systems are being trained using, for example, gender or discrimination bias. This means that the data they output can risk of perpetuating and amplifying existing biases, which may disproportionately impact vulnerable students.

Safely implementing AI in schools and colleges

Given the potential safeguarding risks to students where AI is used in schools and colleges, it is vital that sensible and appropriate implementation of AI is considered.

Some members of staff may already be using AI day-to-day. As a first step, schools and colleges may find it helpful to conduct a consultation with staff to understand how they consider that AI can most helpfully be used. This will also be an opportunity to gather and understand concerns that staff may have in respect of AI; this can be fed into a risk assessment to consider the potential risks of using AI, the rating of those risks, and steps which can be taken by the school or college to mitigate the risk.

Schools and colleges should develop an AI policy, setting out the expectations in respect of use of AI by staff and students. Steps should be taken to ensure that staff understand the policy and receive sufficient training to implement it.

Communications should be provided to staff, but also to students (and parents where appropriate). These communications should clearly set out what AI is, how the school or college will use it more broadly, how it will be used by teachers to support learning, and how students may be allowed to use AI.

Once AI use has been successfully implemented, the school or college will of course have to keep it under review, ensuring to report any incidents where necessary.

If you would like support in respect of implementing AI in your school or college, or review your existing AI policy to ensure that it is robust and appropriately protects your school or college, please get in touch.