AI Use in Educational Psychology: Exploring the Pros and Cons For EPs and Local Authorities

Use of AI is a hot topic across all sorts of professions. Yet, its use in psychology and healthcare services is particularly complex, with a number of potential issues around privacy, safeguarding, ethics and professional standards. That notwithstanding, we cannot afford to ignore the way AI will undoubtedly change the way we all work and operate, and quite possibly, for the better.

AI may enable us to speed up that more mechanical part of the workload, be more productive and therefore, extend our services to even more people. It may enable better ways for EPs, local authorities and other services to work more cohesively together, helping to enhance communications and making processes smoother.

This year, Aubergine teamed up with Educational Psychologist, Maxine Caine, to deliver PsychScope - a series of webinars focused on the legal aspects of educational psychology. There was so much demand from our audience to talk about AI that in episode two, we delivered a jam-packed webinar that delved deep into using AI in an Educational Psychologist Practice.

This article shares some of the most essential information from the session on AI use for EPs and local authorities (LAs) and explores key considerations on AI use in Educational Psychology, as well as ways to mitigate risk.

How AI is Being Used in Educational Psychology

Before we discuss usage, let’s clarify what AI actually is. Artificial Intelligence is autonomous technology capable of not only doing as instructed but also taking actions, making decisions and exploring solutions. Whilst some, such as ChatGPT, Claude and Grammary are obvious AI, we often fail to contemplate how many systems, platforms and software have AI built in now, including search engines, social media, help desks, management systems and much more. AI isn’t just an option anymore; it’s actually increasingly difficult to avoid.

An EP using Grammarly to check their communications or improve documents, or using the Gemini feature on Google as part of their research, or even utilising spam filters for their email accounts, are using AI. As such, we all benefit from knowing a bit more about it and how it works. Before we go into this though, let’s look at some of the ways AI is specifically being used in educational psychology practice:

• Report Writing and Drafting: Some EPs are feeling an immediate time-saving benefit from using tools like Copilot, ChatGPT, Claude, Grammarly, and speech-to-text to assist with:

  
  

  - Structuring reporting

  - Summarising notes

  - Drafting, re-writing and reviewing

  - Generating parent-friendly versions

  - Turning bullet points into prose

  
  

• Psychoeducation and Resources: Things like Canva AI, PowerPoint Copilot, and safe image generation tools help create worksheets, visual explainers, social stories, and child-friendly materials.

  
  

• Practice-Management Systems (with built-in AI): We have tools like WriteUpp, Cliniko, Semble and many of these systems are quietly adding AI features behind the scenes, like autosummaries, suggested next steps, and template support.

  
  

• Productivity and Workflow: AI email drafting and scheduling assistants, like Notion AI, Zapier AI — are automating the invisible admin that eats into time.

  
  

• Research and Evidence Scanning: For research and evidence scanning, there are tools such as Semantic Scholar AI, Elicit, Copilot for PDFs and these tools can summarise research papers, extract key findings, and help EPs stay on top of evidence.

  
  

• Emerging Agentic Tools: Agentic AI is a different beast. These tools are more than chatbots. They’re capable of running tasks and generating materials with limited supervision. Learning from themselves, they break down the main objective into a series of smaller tasks, using a variety of tools to accomplish the end goal. For example, when you instruct Claude to put together a case file it will utilise what it already knows and what has been input to draft this, going through a series of steps to pull all the necessary information and repurpose that to fit the bill. Or, Copilot, for instance, can use one workflow to produce a parent-friendly summary from consultation notes. These are examples of agentic AI and this is really reshaping how we work. However, the moment an AI system starts ‘acting’ rather than just ‘answering’, the risks increase but your professional responsibility does not.

In the next section, we’ll flag some potential risk from using both agentic and non-agentic AI, helping you identify when use of AI could make you non-compliant with existing regulations, or even cause harm to your practice or your clients.

Key AI Considerations for Educational Psychology

Educational Psychologists operate in one of the most complex professional ecosystems out there. As well as dealing directly with children/young people and their families, you may also be working with schools, local authorities, health professionals, social care and sometimes courts and tribunals. So, it’s not just about whether you are using AI. When working with other organisations, you’ll also need to know about and consider their application of AI.

This is where the AI issue gets more difficult. However, really it’s just an extension of the other legal considerations you’ve likely had to manage when sharing information with other bodies, such as GDPR, consent and privacy rights.

Now that AI is woven into so many of the tools we already use, there’s a real risk of soliciting those tools without fully understanding the new risks and potential implications.

Use of AI can lead to breaches of -

• Data protection law - especially around children’s data and mental health data, which are the highest risk categories under UK GDPR.

  
  

• Professional standards - set by BPS, HCPC and the ICO.

• Safeguarding obligations  - because AI can distort meaning, remove nuance, or misrepresent risk.

So, whilst AI can certainly be beneficial in the Educational Psychology profession, we must endeavour to understand how it works so we may use it safely and responsibly. It’s essential to know that Agentic AI doesn’t remove your responsibility just because it’s acting by itself. You are still in the driver’s seat, so to speak, and therefore still responsible.

What UK Law States About AI

The UK has yet to draft any AI-specific legislation. Unlike the EU, which entered an EU AI Act into force in August 2024, the UK is using existing laws and regulations, overseen by bodies such as the ICO, CMA, Ofcom and MHRA.

So UK GDPR reigns supreme for data protection and although AIs are subject to uphold these regulations, not all are designed for use in the UK, which means it’s necessary to check that their policies uphold UK GDPR before using them.

Be selective in the tools you use and take time to understand how they work, being careful to ensure that no data you input is ever going to be made public.

Issues With Using AI In Educational Psychology:

• Privacy

  Some Open AI tools, which often allow for free access, harness data for training purposes, and this can pose a risk that anything shared with the AI tool could be used or shared elsewhere. Make sure you’ve done your due diligence and aren’t using anything that could compromise client privacy.

• Accuracy

  As a research tool, AI is beneficial but not completely reliable. Whilst many of us check sources and only visit reliable outlets for information, the hype around AI has made some of us somewhat blind to trusting the information it provides. Be careful, as it can be so confidently wrong sometimes!  Always request sources and check source material, as AI is programmed to please and prone to hallucinating.

• AI Bias

  AI mines existing data as a source of information, and therefore, there is a real risk of AI reinforcing cultural, linguistic, or diagnostic biases, standards, especially in SEND contexts. The nuance that you have developed in your career may not yet exist in AI tools, so it’s important to screen, review and be ready to amend anything that is not inclusive. Especially that which may breach UK discrimination laws, governed by the Equality Act 2010.

• Professional Accountability

  As impressive as some of these tools are, don’t lose sight of the fact that you are the expert. As an Educational Psychologist, you’ve been trained, have had experience and are governed by bodies such as the BPS, HCPC and the ICO. Whilst AI is supposed to adhere to these standards, that might not always be the reality, so make sure you are fully schooled in the regulations prior to seeking AI assistance and then apply your professional oversight and scrutiny to anything produced or enacted.

• Consent and Transparency

  Just as you must obtain consent for any collection of data and permission to work with anyone, particularly minors, you must also be transparent with your use of AI tools. It’s essential to update your policies and contracts to explain both the risks and how you mitigate them before you engage in any working relationship where AI tools may be utilised.

  The most important thing is to be transparent. This protects your business but also can awaken us to risk, leading us to be better prepared to mitigate that.

• Environmental Impact of AI

  Often overlooked is the environmental impact of AI use. A single generative AI query can consume 4 to 5 times the energy of a standard web search, so if you’re seeking to lessen your carbon footprint, then you should use AI tools conservatively and consider the environmental cost.

Managing and Mitigating AI Risk

Working in Educational Psychology, you’re likely aware of the importance of DPIAs (data protection impact assessments). You probably already use these. If you’re using AI though, or even if other organisations you’re working with are using AI, then this/these may need updating.

DPIAs are about identifying potential risks and safeguarding against them, and are a legal obligation under UK GDPR, as well as a professional responsibility.

Your DPIA may need updating to identify AI risk specifically and set out how you intend to prevent or minimise this. You might also include how you intend to manage data protection when working with other bodies that may be harnessing AI in their workflow.

One measure you might consider is placing disclosures on certain documents. For instance, stating on a report that it must not be uploaded into AI tools (e.g., ChatGPT, Gemini, Copilot, Claude) and emphasising that doing so may breach intellectual property rights and may expose the child’s sensitive data to third‑party systems.

There are also a number of documents you may need to update if using AI in your practice, including:

• Client Terms and Conditions

• Privacy Policy

• Special Category Data Consents

• Data Protection Impact Assessment

You might also consider having an AI Policy in place and providing child-friendly AI explainers to young people.

Legal Resources for Educational Psychologists

Ensure your EP practice stays compliant when integrating AI with my essential FREE AI Compliance Checklist for Educational Psychologists.

This checklist gives you a list of things to consider and will help you navigate legal, ethical, and data protection requirements when using AI tools in your practice and across multi-disciplinary teams. There are also sections on safeguarding issues and things to think about when using AI within schools.

For a more comprehensive legal pack, designed specifically for Educational Psychologists, download my Legal Toolkit for Educational Psychologists. This contains template agreements, policy templates, website documents, consents, information sheets for pupils, UK GDPR Guidance, Safeguarding and Child Consent Guidance, as well as all the documents you need if you're using AI.