As the European Union takes steps to regulate artificial intelligence with its AI Act, there’s a lot in here that UK schools should be mindful of. While we’re no longer bound by EU regulations, the Act offers many valuable insights into the future of AI governance in education and there’s lots to like. Here are my top ten key implications for UK schools:
1. Rethinking AI Risk Assessment
The EU AI Act introduces a risk-based approach to AI regulation. I mean, we should be anyway, but for UK schools, if you’re not already, you should start thinking critically about the AI systems you use. Any AI used for determining school admissions, evaluating learning outcomes, or monitoring student behaviour during exams would be considered high-risk under the EU framework.
What does that mean for UK schools?
Conduct a thorough audit of all AI systems in use and assess their potential impact on students. Create an AI inventory, listing all AI tools used in your school, their purpose, and potential risks. Involve teachers, IT staff, and administrators in this process to ensure a comprehensive overview.
2. Enhanced Data Protection Measures
While UK schools already adhere to UK GDPR, the EU AI Act emphasises even stronger safeguards for AI systems processing children’s data. This signals a trend towards more stringent data protection in AI-driven educational tools.
What does that mean for UK schools?
Review your data protection policies to ensure they cover AI-specific concerns such as algorithmic bias and automated decision-making. Conduct a fresh Data Protection Impact Assessment (DPIA) (and maybe update your existing DPIA scheme to include this) for each AI system, particularly those handling sensitive student data. Your Data Protection Officer should be front and centre in this process.
3. Transparency in AI Decision-Making
The EU Act places a strong emphasis on transparency in AI systems. This trend is likely to influence UK practices as well, pushing for more openness about how AI makes decisions in educational contexts. Hopefully, vendors who adopt a ‘privacy by design’ approach will make this easy for you, but that may just be wishful thinking.
What does that mean for UK schools?
Prepare to be able to explain how AI systems make decisions, particularly those affecting students’ educational journeys. Consider developing an ‘AI in Our School’ guide for students and parents. Use simple language and infographics to explain how AI is used and how it impacts learning and outcomes.
4. Regular AI Impact Assessments
Under the EU framework, high-risk AI systems require ongoing monitoring and assessment. This approach could become standard practice in the UK as well.
What does that mean for UK schools?
Consider implementing a protocol for regular risk assessments of your AI tools. Establish a termly ‘AI impact review’ meeting involving key stakeholders. It might seem like overkill but I would suggest gathering feedback from your AI working party (already likely to contain teachers, students, and parents) about the effects of AI tools on learning and well-being.
5. Prioritising AI Literacy
The EU Act indirectly promotes AI literacy, a crucial skill for the future. Regardless, it would be my strong recommendation that all schools should prioritise this area to ensure students and all staff are well-prepared for an AI-driven world.
What does that mean for UK schools?
Developing a comprehensive AI literacy programme for both students and staff is a no-brainer. Of course, to help with effective training you could consider engaging experts like myself, either with the work I’m undertaking in collaboration with InnerDrive offering Inset days, or at a bespoke level, as I have been for many years, just drop me a line via my contact page if you’re interested. My expertise can help focus on how to use AI effectively to support learning, teaching and workload reduction for teachers and in a similar fashion with support staff, all while being mindful of ethics and bias. This approach ensures that your school community is not only familiar with AI but also equipped to use it responsibly.
6. Establishing AI Governance Structures
The Act emphasises clear accountability in AI deployment. UK schools should consider similar governance structures to ensure responsible AI use.
What does that mean for UK schools?
Consider establishing an AI governance committee within your school including representatives from teaching staff, IT, senior leadership, students, the governing body and parents. Meet at least termly to discuss AI implementation and address any concerns.
7. Ensuring Fairness and Non-Discrimination in AI Systems
A key aim of the EU AI Act is to prevent AI from perpetuating or exacerbating discrimination. This focus on fairness should be a priority for UK schools as well.
What does that mean for UK schools?
Rigorously test any AI systems for bias before implementation, particularly those used in admissions or assessment processes. Perhaps adding this as an additional element to your DPIA checks before adoption would be a prudent measure, but certainly, conducting a ‘fairness audit’ of your existing AI systems would be helpful. Analyse outcomes across different student groups to identify any disparities that might indicate bias.
8. Preparing for Future UK AI Regulations
While the UK isn’t bound by the EU AI Act, similar regulations will likely be introduced here in the future. Preparing now will put UK schools ahead of the curve.
What does that mean for UK schools?
Stay informed about AI policy developments. The DfE have been pretty good on this already and with Bridget Phillipson’s fantastic start to her tenure, I would hope it would continue to do so. Assigning a member of staff to monitor UK AI policy discussions and report back to the school leadership team regularly would be a prudent measure, such as your teaching and learning or curriculum leads. For a solid starting point in developing your school’s AI policy, consider using the AI policy that I developed which I will be updating soon with version 2, given the measures in the EU AI Act and the advice I’m sharing here. The resource will help give you a leg-up and provide a framework to get you started navigating the complexities of AI implementation in schools.
9. Proactive Engagement with AI Vendors
The EU Act places significant responsibility on AI providers to ensure their systems are compliant. UK schools should adopt a similar proactive approach with their AI vendors.
What does that mean for UK schools?
Engage proactively with your AI vendors. Develop a standard ‘AI vendor questionnaire’ covering key areas of compliance, data protection, and ethical AI use and use this when considering new AI tools or when reviewing existing ones.
10. Fostering an Ethical AI Culture
The EU AI Act promotes a culture of responsible AI use. UK schools should strive to create an environment where ethical considerations are at the forefront of all AI-related decisions which I would suggest should also include thinking around sustainability. AI doesn’t have to be used for everything. As I’ve shared before, AI can sometimes be a little like using a sledgehammer to tap in a nail, so (linking back to AI literacy) ensuring colleagues are trained and supported with CPD so they can make cognisant choices about use will help ensure the choice of tool aligns to the job at hand.
What does that mean for UK schools?
Consider integrating AI ethics into your school’s existing ethics or PSHE curriculum. Encourage students to debate the ethical implications of AI in education and society at large, perhaps as part of a new digital cognition curriculum, such as which I’ve been developing with some of my client schools, including AI literacy. It should go without saying that ethical considerations should form a key part of any decision to implement new AI tools.
Looking Ahead
Whilst the EU AI Act doesn’t directly apply to UK schools, its principles help us with guidance on what we can expect in the future here in the UK and help us navigate these new, uncharted digital waters. I hope that by considering the points here you can ensure you’re using AI responsibly, ethically, and effectively.
Remember, if you want help, please do get in touch, but in the meantime, the EU AI Act provides a valuable roadmap, and hopefully some of these pointers will help you prepare for what’s coming in the next academic year.
