What every educator, school and vibe coder should know about data protection.
With the rise of vibe coding, teachers can now create sophisticated educational applications in a matter of hours. But whilst AI has made software development considerably easier, our responsibilities when handling people’s data have not disappeared.
Having recently worked through a privacy and security review of my own JSFun educational tools, I thought it would be useful to explore what GDPR compliance actually means and, perhaps more importantly, what it doesn’t mean. (Final privacy notice here.)
What is GDPR?
The General Data Protection Regulation (GDPR) is European legislation designed to protect people’s personal information. The UK has its own version, UK GDPR, whilst Malaysia has the Personal Data Protection Act (PDPA).
Although these laws differ, they share an important principle: organisations should handle people’s information responsibly.
GDPR is not simply about passwords, encryption or cookie banners. It is about the entire lifecycle of personal information, from the moment it is collected to the moment it is deleted.
The seven principles of GDPR
| Principle | What it means in practice |
|---|---|
| Lawfulness, fairness and transparency | Be honest about what you do with data and have a lawful reason for using it. |
| Purpose limitation | Use information for its intended purpose. |
| Data minimisation | Only collect what you actually need. |
| Accuracy | Keep information accurate and up to date. |
| Storage limitation | Do not keep information indefinitely without justification. |
| Integrity and confidentiality | Keep information secure. |
| Accountability | Be able to demonstrate responsible data handling. |
For educators developing their own applications, data minimisation is particularly important. Does your spelling quiz really need a student’s email address? Does your maths game need their date of birth? Does your classroom timer need a Google login? Every additional piece of personal information brings additional responsibilities.
Ten myths about GDPR
| Myth | Reality |
|---|---|
| Every website needs a cookie banner. | It depends on the technologies used and applicable privacy laws. |
| You always need consent. | Consent is one of several lawful bases for processing information. |
| You cannot collect personal information. | You can, provided there is a lawful basis and appropriate protection. |
| All data must be stored in Europe. | International transfers are possible with appropriate safeguards. |
| Removing names makes data anonymous. | Other information may still identify individuals. |
| A privacy notice makes you compliant. | Your actual practices must match what the notice says. |
| Using Google or Microsoft makes you compliant. | You remain responsible for how you use their services. |
| Free educational applications are exempt. | Being free does not automatically exempt an application. |
| GDPR compliance means an application is safe for children. | Safeguarding and content safety are separate considerations. |
| GDPR compliance requires a certificate. | GDPR certification is voluntary. |
Try a matching puzzle>> (Useful for professional development)
Privacy by design
Imagine that we are developing a simple classroom quiz. There are two possible approaches.
Traditional cloud application
- Students create accounts.
- Names and emails are collected.
- Answers are sent to a server.
- Results are stored centrally.
- Individual activity is recorded.Greater privacy responsibilities
Privacy-first application
- No accounts required.
- No names or emails collected.
- Answers are processed locally.
- Results remain on the device.
- No individual tracking.Reduced privacy risk
Both applications could provide exactly the same learning experience.
Of course, there are legitimate reasons for storing results centrally, such as monitoring student progress. However, we should make that decision deliberately rather than simply collecting information because the technology allows us to.
The easiest data to protect is the data you never collect.
The hidden data problem
Even if your application never asks for a student’s name, information may still be transmitted through:
- Website analytics.
- Externally hosted fonts.
- JavaScript libraries.
- Embedded YouTube videos.
- AI services.
- External image requests.
- Website hosting logs.
- Information contained in shared URLs.
An AI coding assistant might introduce some of these dependencies without you realising it.
When reviewing my own JSFun website, it was important to look beyond what the user could see on the screen and examine what the underlying code was actually doing.
Privacy is a technical consideration as well as an administrative one.
Can AI help with GDPR compliance?
Absolutely. AI coding assistants can help developers:
- Identify external network requests.
- Detect cookies and browser storage.
- Find potentially insecure code.
- Replace external dependencies with local alternatives.
- Create privacy notices.
- Document data processing.
- Implement privacy settings.
However, an AI-generated privacy notice or an automated code review is not proof of legal compliance. Human oversight remains essential, particularly when determining lawful bases, contractual arrangements and safeguarding requirements.
Five questions to ask before publishing an application
Pre-publication privacy checklist
- Do I actually need to collect this information?
- Where does the information go?
- Who has access to it?
- How long will it be retained?
- Could I provide the same functionality without collecting it?
Conclusion
GDPR compliance is not the same as safeguarding. An application could handle personal information responsibly whilst still exposing children to inappropriate content, unrestricted communication or unsuitable AI-generated responses. Equally, an excellent educational resource could have poor data protection practices. Schools must consider both privacy and safeguarding when evaluating digital tools.
GDPR should not be seen as a barrier to educational innovation, but as an opportunity to build better applications. As vibe coding makes software development accessible to more educators, privacy by design needs to become a fundamental part of digital literacy. We should not simply ask whether we can build something, but whether we can build it responsibly. Build useful tools. Collect less data. Protect what matters.
