Written evidence by Katie Brockhurst (AIE0049)

Education Committee

The use of Artificial Intelligence and EdTech in Education

I am Katie Brockhurst, a digital wellbeing specialist, educator and author with over two decades of experience working inside digital ecosystems. I worked as a Social Media Manager at Director level before moving into advocacy and education, working directly with schools, teachers and young people.

I created The Attention Experiment, a schools programme designed to help young people understand how digital platforms are engineered to capture and hold their attention, and to develop the agency to make more intentional choices about how they engage with social media specifically.

I am also the author of the Social Media for a New Age book series, a Business Book Awards finalist, gold Sony Award winning Podcaster and hold a Diploma in Digital Wellbeing from the Consciously Digital Institute, with certification from the Europass Teachers Academy.

I am currently resitting the Consciously Digital Diploma in 2026, having originally graduated in 2020, in order to stay current with the rapidly evolving research on AI, cognitive impact and digital behaviour. The pace of change in this field means that even qualified practitioners must continually update their understanding — which underlines how far behind the general population, including teachers and parents, currently are.

I have delivered sessions in schools and sixth form colleges including Sherborne School and Strode College, and I’m currently working with EducationScape — the organisation behind the Festival of Education and Schools Week — and the School Speakers agency to deliver digital wellbeing sessions to students, teachers and parents.

Having witnessed, experienced and studied first-hand how addictive design was built into digital platforms, I bring this perspective with the hope that it helps your inquiry

Challenges and Opportunities

One of the most important things I have learned is that our relationship with technology is personal, subjective and unique. This is as true of AI as it is of social media, and it has profound implications for how we think about deploying it in educational settings.

I am not opposed to AI in education. AI tools have genuine potential for accessibility, for reducing teacher workload, for supporting students who learn differently, and as a thinking partner when used with intention and awareness. My own use of AI in preparing this submission is an example of that. But these benefits are only realised when the tools are deployed with an understanding of how they work, what they ask or don’t ask of the user, and what the risks are when that understanding is absent.

When I talk with both young people and adults about AI, I see a mix of confusion, excitement, fear, frustration and genuine curiosity. What I do not see is adequate education or understanding of the mechanics and business models of many of these technologies and how this may affect our mind, body and overall wellbeing.

We need to understand these tools ourselves so we can teach young people what AI is, how it affects the brain and body, and what it means to be growing up in a world where AI technologies are now embedded into daily life.

I do not see young people being taught enough about cognitive offloading, the neuroscience of attention, or the short and long-term wellbeing impacts of outsourcing their thinking.

Nor are they being helped to fully understand what is happening when they talk to a chatbot, or why we have increasingly complex relationships with these technologies.

This urgently needs addressing.

I bring these ethical concerns from working inside the social media industry, where an architecture of digital addiction has been built into the algorithms as part of their business models. This was clear: capture attention, extract data, monetise both. Our wellbeing was not a design priority. It has taken the better part of a decade for the public conversation to catch up with this, despite many people trying to highlight it. The recent US court case in which a jury found Meta and YouTube negligent for addictive design marks a significant shift in accountability.

Yet several of the companies moving products into educational settings are the same companies whose practices around child safety and data ethics have been the subject of legal action, parliamentary inquiry and regulatory scrutiny. This should give the Committee cause for concern.

An ethical framework cannot be built on the assumption that these companies will self-regulate or adjust to meet the best outcomes for children. The history does not support that assumption.

These concerns are not just about corporate behaviour, but the impact of this we see playing out in our classrooms and homes every day.

When I delivered The Attention Experiment at Sherborne School, the school told me that one of their biggest challenges was that students could not differentiate between how they use their Microsoft Surface device for school work and how they use their personal devices. There was no difference for them. The same habits were present across devices. Students were checking emails, installing games and getting pulled into distractions on the same device they were expected to do focused work on.

I taught a session to Year 10 boys on what happens in the brain when you focus on a task without interruption, compared to what happens when you break that focus to check an email or switch between apps. This is the kind of foundational understanding that is missing, not just for students, but for the schools deploying these digital tools.

Research consistently shows that reading on paper produces better comprehension than reading the same material on screens, and that digital distractions in learning environments have a measurable negative effect. If schools are increasingly delivering learning through connected digital devices and integrating AI tools into the classroom, these findings should inform how and when those technologies are used, not to abandon them altogether, but as evidence that using them in classrooms requires far more thought than it is currently receiving.

Onto your specific question of deskilling: Research from MIT Media Lab (Kosmyna et al., 2025) found that students using ChatGPT to complete essay writing tasks showed significantly reduced neural connectivity compared to those working independently, with their brain engagement dropping by over half. When those students were subsequently asked to work without AI, their brains continued to underperform compared to students who had never used AI at all. The majority could not recall a single sentence from essays they had written minutes earlier. The researchers coined the term “cognitive debt” to describe how AI spares mental effort in the short term but generates long-term costs.

It is worth noting that this study does not appear to explore the interesting middle ground, of what happens when AI is used to support and deepen thinking rather than simply replace it. I am aware of the work by Dr Sam Illingworth, Full Professor of Creative Pedagogies at Edinburgh Napier University, founder of Slow AI, a curriculum for critical AI literacy, who has given expert evidence to the House of Lords Select Committee on AI literacy.

The Impact on Learning

Through regular calls with the Consciously Digital Institute, I have access to the latest peer-reviewed research on digital behaviour and cognition, and have encountered a growing body of evidence that raises serious questions about the long-term impact of AI tools on the essential skills this inquiry is asking about, specifically around critical thinking, oracy, creativity and problem solving.

Research by Gerlich (2025) found a significant negative correlation between frequent AI tool usage and critical thinking skills. The study identified cognitive offloading as the key mechanism, and showed that younger users aged 17–25 were the most susceptible. The author concluded that while AI tools are designed to free up cognitive resources for more complex thinking, in practice the opposite is occurring, that instead they are leading to a decline in cognitive engagement and skill development.

Understanding why this is happening requires looking beyond the tools themselves to the business models behind them.

Dr Anastasia Dedyukhina recently presented insight on the evolution of digital business models regarding AI which I also think is important to add to this enquiry.

She describes a shift through three economic stages: the attention economy, where platforms capture what you see; the intention economy, where AI systems capture what you want; and the affection economy, where AI captures how you feel, with AI models ever hungry for more input to train on.

This matters for the Committee’s inquiry because these are not future concerns. Young people are already forming relationships with AI chatbots, which can begin with students simply asking for help with their homework.

They are already disclosing needs to systems designed to maximise engagement through empathy cues, emotional mirroring and relational continuity. And they are doing so largely without any educational framework to help them understand why these interactions feel meaningful, what is being extracted from them, or how these systems are designed.

The question of digital literacy education which the Committee raises, cannot be answered adequately without this context. Teaching young people to use AI “safely and responsibly” is insufficient if it does not include an understanding of the economic models driving AI design, the neuroscience of attention and reward, and the difference between a tool that supports your thinking and one designed to capture emotional investment.

This is the work I am doing through The Attention Experiment, by helping young people understand the mechanisms of how apps capture their attention, using this attention sequence; Scroll → Loop → Notice → Choose, that explains what's happening in the brain at each stage, why young people's developing brains are especially impacted, and why the latest research shows that the best approaches are body-based, not just information-based.

But this work needs to be embedded systematically across the education system, not left to individual practitioners, especially where AI tools are being used in schools and by students, which we know they are already. Parents, too, lack access to this important information that leaves us all vulnerable to big tech.

If digital wellbeing educators are only now beginning to understand the implications of digital technologies and AI across society and education, parents are further behind still, and the children navigating these systems are the least equipped of all.

My recommendation to the Committee is that digital literacy education must go beyond teaching young people how to use AI tools. It must equip them with the agency to understand and recognise when tools are designed to keep them online, what is being asked of them when they engage, and how to make intentional choices about that engagement.

The education system must help them develop agency and awareness, which requires embedding the neuroscience of attention, the economics of platform design, and the principles of critical thinking into the curriculum, not as ‘specialist’ assemblies and guest lecture talks but as foundational knowledge for navigating the world they are growing up in.

Without this, we are not preparing young people for an AI-shaped future, because we have not been prepared adequately as it is, let alone for what is coming next.

Children’s Digital Rights

I've spent my career working inside social media platforms, and saw how addictive design was built and deployed into what millions of children now use every day and struggle to put down. I watched it happen and have tried to raise awareness. When The Social Dilemma came out on Netflix in 2020, addictive by design became mainstream knowledge, yet it has still taken this long for anyone to be held accountable for the harm it has caused and for governments to finally take action.

The Committee asks about the evolving nature of digital risks, including addictive design features and the impact of generative AI. From my experience, the risk with AI is not separate from the risk we saw with social media, which I see as an extension. The same design logic that kept children scrolling is now being applied to AI tools currently available in the public domain, that respond to them personally, learn from their behaviour, and are designed to deepen engagement over time. The difference is that these interactions are more intimate — a chatbot that mirrors your emotions and remembers your conversations creates a very different kind of relationship than a personalised newsfeed.

The safeguarding implications of this are significant. Even where AI tools in education are not designed as companions, they may still be built to maximise engagement, they may still collect and monetise student data, and the cognitive wellbeing of the child may not be a design priority. These are safeguarding concerns that current frameworks may not be equipped to fully address.

The Committee also asks how children can be protected from online harm while their rights to participate and access information are upheld. I would argue that genuine protection cannot come from restriction alone — it must include education. A child who understands why a chatbot feels like a friend, why an algorithm shows them what it shows them, and what is being collected when they engage, is better protected than one who is simply told not to use it. This is the case for the attention literacy and cognitive awareness I have recommended embedding into the curriculum.

We cannot afford another decade-long delay before we start helping students understand these technologies.

Summary of Recommendations

Based on the evidence and experience outlined in this submission, I recommend that the Committee consider the following:

1. Embed attention literacy and cognitive awareness into the curriculum

Embed attention literacy, cognitive awareness, and an understanding of how digital platforms are designed into the curriculum as foundational knowledge. The 2025 Curriculum and Assessment Review has rightly introduced media literacy and digital safety, but this does not yet extend to teaching young people about the neuroscience of attention, the business models driving platform design, or the cognitive impacts of outsourcing thinking to AI. These should be part of core provision, not specialist assemblies or guest lectures.

2. Strengthen teacher training and CPD on AI and digital systems

In my experience working with teachers who are interested in digital wellbeing, many are already overstretched. They recognise its importance and are witnessing significant changes in how their students think and operate. However, they often struggle to manage their own attention and relationship with technology, and lack a clear understanding of what digital and AI technologies are, how they work, why they are designed in this way, and the impact this has.

I would recommend that teacher training and CPD on AI address not only how to use these tools, but also the business models, design ethics and cognitive impacts behind them — so that educators can make informed decisions and guide students in using these technologies critically and responsibly.

3. Establish clear boundaries between learning technologies and distraction ecosystems

Provide clear guidance for schools on managing the boundaries between learning technologies and distraction ecosystems. A connected device used for schoolwork is also a gateway to email, games, social media and AI chatbots, all of which have well-documented effects on attention, focus and learning outcomes. Schools should be supported in creating environments that protect sustained attention and minimise cognitive fragmentation.

A note on the preparation of this submission

This submission was prepared with the assistance of AI tool, Claude. I have used it in line with an ethical framework for working with AI, developed by the Consciously Digital Institute, which distinguishes between AI as a substitute for thinking and AI as a support for it.

I used it as a research organiser and drafting collaborator, while all evidence, arguments, perspectives and editorial judgements in this document are my own.

I am flagging this not only for transparency, but as an approach that is intentional and values-led.

May 2026