Artificial Intelligence offers many opportunities for education and it is important to use suppliers committed to transparency, security, and ethical practices.
At Welbee, we’ve built Edu Intelligence on a foundation of trust, security, and a deep understanding of Education and the needs of school, trust and group leaders — so you can have full confidence that our powerful education AI is safe, ethical and secure. We also want to help you make sure that any other suppliers you consider or use are secure, so we have put together 9 key questions you should ask any supplier who processes data or uses AI in their product.
Why It’s Important: This is about protecting personal privacy and complying with GDPR and the DPA 2018. Many AI systems analyse data by accessing raw, identifiable information – for instance, being sent a lot of information containing specific details of students, such as “Tom Smith in Year 9 is eligible for Pupil Premium” and allowing the AI to analyse this, together with the other individual data and identifiable characteristics to form a picture of what is happening. This presents significant data privacy concerns and could lead to complaints from families, students and staff.
The more secure approach is for AI to access only aggregated and anonymous data, which will still yield deep insights without compromising individual privacy.
What to Look For: A clear answer and explanation of what data the AI can access and interact with. If it does use identifiable data, how is this protected? You should only use AI that provides analysis and insights without needing to process information that can identify individuals.
The Edu Intelligence Approach: We’ve designed Edu Intelligence so our AI delivers powerful analysis and tailored suggestions without ever having access to individual student, family, or staff data. We achieve this by ensuring our AI works with aggregated and anonymised information, allowing us to provide deep insights while upholding the privacy of your stakeholders and community.
Why It’s Important: This prevents the chance of your data becoming unintentionally mixed with other organisations data — especially when it’s being accessed by AI.
True separation means it’s in its own dedicated dataset or database, not just part of a larger, shared database where your data is separated only by a “customer ID” or “school ID.”
What to Look For: Assurance that your data is kept in its own secure and separate “container,” database, or dataset, preventing cross-contamination or unauthorised access from other clients.
The Edu Intelligence Approach: Your institution’s data is stored separately in your own secure data warehouse. We ensure this isn’t just a label in a shared system; it’s a structurally separate environment, which is a core part of our security promise to you.
Why It’s Important: These aren’t just guidelines; they’re legal requirements for protecting personal data, especially for children. A supplier should prove they take these laws seriously to protect your institution from privacy and legal risks, particularly when your data is being processed or AI is involved.
What to Look For: Evidence of robust data protection policies, staff training, and processes that are regularly reviewed and updated. Ask about their Data Protection Officer (DPO), how they handle data subject access requests, and specifically how their AI design aligns with these regulations.
The Edu Intelligence Approach: Compliance is built into the DNA of Edu Intelligence. Our system’s design, which limits AI access to identifiable data and ensures separate data storage, is in full compliance with GDPR and other data protection principles.
For instance, by not having our AI process identifiable data, we minimise risks associated with automated decision-making on individuals. We are committed to ethical practices and meeting all regulatory requirements, designing our services with the best interests of all users, including children, at the forefront.
Why It’s Important: Your data is yours. It should be used to benefit your school, trust or group. You need to be sure it’s not being used to build or improve products or underlying AI processes for other paying customers, or for the supplier’s general research and development, without your explicit consent and clear benefit.
What to Look For: A clear “no,” or a very transparent explanation and agreement if it is used (and how you benefit). Your data should not be a free resource for their wider product or process development.
The Edu Intelligence Approach: Your institution’s data.is not used to train AI models or to generally improve our AI processes for other clients or our future products outside of the services we provide directly to you. The insights generated and any system learning from your data are for your institution alone.
Why It’s Important: The thoughts, feedback, and ideas shared by your school community in surveys or other data inputs are valuable. This IP should remain yours and not be used by the supplier for their own separate commercial ventures. This is also a key requirement from the DfE’s guidance on generative AI in Education.
What to Look For: Clear terms stating that you retain ownership of your data and any IP generated from it. The supplier should not have rights to use this content to train their general models or for other commercial uses outside of the service provided to you.
The Edu Intelligence Approach: With Your Data, Your Control as a core principle for us, any intellectual property within your data remains under your institution’s stewardship. Since your data isn’t used for general AI model training, the unique insights and content from your staff and students are protected from any such external use by us.
Why It’s Important: “Privacy by Design” means the system was built from the ground up with data protection in mind, rather than having privacy features bolted on later. This includes considerations like physical data separation where appropriate, ensuring that data is only collected for specified purposes, and that access is strictly controlled. This proactive approach is more secure and reliable.
What to Look For: The supplier should be able to explain how the system design incorporates these principles, such as collecting only necessary data, using it only for agreed-upon purposes, implementing strong access controls, and ensuring data is stored in secure, appropriately separated environments.
The Edu Intelligence from Welbee Approach: We built Edu Intelligence with a privacy first approach. This includes how we handle data storage, ensuring your data is in its own secure data warehouse. Our AI provides insights without needing access to identifiable individual data, inherently minimising data exposure. We focus on collecting only the data needed to deliver actionable intelligence for your educational goals.
Why It’s Important: AI models can inadvertently learn and perpetuate biases present in data, leading to unfair or discriminatory outcomes. Ethical AI design and auditing are crucial for fairness.
What to Look For: Information on their ethical AI framework, how they test for and mitigate bias, diversity in their development teams and data sets (if applicable to the AI’s function), and transparency about the AI’s limitations.
The Edu Intelligence Approach We are committed to Ethical and Secure AI. Our system employs multiple AI ‘agents’ that work without accessing raw identifiable data, focusing on objective analysis to provide evidence-backed suggested actions.
Why It’s Important: Education isn’t just another business sector. It has its own language, metrics, and complex student groups. A generic AI might not understand school-specific data, challenges (like attendance or staff wellbeing), or be able to make meaningful comparisons like “how do SEND PP Y8 girls compare to SEND PP Y9 Boys?”
What to Look For: Proof that the AI understands educational contexts, terminology (like SEND, Pupil Premium, KS2/3/4, etc.), and goals. Look for case studies or features developed with input from educators and an ability for the AI to interpret and report on specific educational cohorts.
The Edu Intelligence from Welbee Approach: Edu Intelligence is built for education by educators. We’ve designed it to understand the specific language and complex questions relevant to schools. So, when you need to analyse “how SEND PP Y8 girls compare to SEND PP Y9 boys” in terms of wellbeing or attendance, our system is built to provide those specific insights. It’s designed to connect diverse data points like attendance records with stakeholder feedback on wellbeing and school trends, addressing core educational challenges. Our focus is entirely on helping schools, trusts, and groups thrive.
Why It’s Important: If an AI tells you to do something, you should be able to understand why. “Black box” AI, where the reasoning is hidden, makes it hard to trust the advice or explain it to others.
What to Look For: Suppliers who can clearly explain how their AI arrives at conclusions, the key data points it considers, and the evidence base for its recommendations. Avoid systems where insights are presented without justification.
The Edu Intelligence Approach: We aim to remove guesswork with clear insights, identified priorities, and practical, evidence-backed suggested actions — and our AI will show information that explains why it has identified insights and made particular suggestions.
Discover how Edu Intelligence from Welbee will help your school, trust, or group harness the power of data and AI responsibly and effectively to help you drive school improvement.
