Stop Uploading School Data to ChatGPT Until You Know This
リアクション
2026年08月29日
AI is already being used across schools — but do we actually understand what happens to school data when we put it into AI?
In this video, I look at one of the most important questions for teachers and school leaders using ChatGPT, Gemini, Claude and other generative AI tools:
Where does your data actually go?
Uploading a SEF, school improvement plan, assessment information or other internal document isn't simply a question of whether “AI is safe”.
The type of account you're using, the information contained in the document, the provider's terms, retention, training policies, organisational controls and the architecture surrounding the AI all matter.
And this is where we need to move beyond simply telling teachers which AI tools they can and can't use.
For me, this is part of what it means to Lead with AI.
Schools need to understand:
→ What data are we sharing?
→ Where is it going?
→ What happens to it?
→ Who controls it?
The distinction between a personal AI account, an organisation-managed education/business environment and a purpose-built application using an API is also hugely important. The underlying AI model might even be the same — while the data journey surrounding it can be very different.
Current UK Department for Education guidance specifically asks schools to consider how generative AI tools use personal data and recommends using approved tools, involving data-protection leads and avoiding personal data unless its use is necessary and properly protected.
This distinction is becoming increasingly important as education-specific AI environments emerge. For example, OpenAI states that information shared within ChatGPT for Teachers is not used to train its models by default and provides school/district administrative and privacy controls.
DEEP AI Audit
This is also exactly why we're building the DEEP AI Audit.
An AI policy tells you what should be happening.
An AI audit helps you understand what is actually happening.
Which AI tools are being used? By whom? For what purposes? What information is being shared? And where are the potential gaps in governance, knowledge and practice?
DEEP AI Audit: https://audit.deepeducationnetwork.com/
In this video, I look at one of the most important questions for teachers and school leaders using ChatGPT, Gemini, Claude and other generative AI tools:
Where does your data actually go?
Uploading a SEF, school improvement plan, assessment information or other internal document isn't simply a question of whether “AI is safe”.
The type of account you're using, the information contained in the document, the provider's terms, retention, training policies, organisational controls and the architecture surrounding the AI all matter.
And this is where we need to move beyond simply telling teachers which AI tools they can and can't use.
For me, this is part of what it means to Lead with AI.
Schools need to understand:
→ What data are we sharing?
→ Where is it going?
→ What happens to it?
→ Who controls it?
The distinction between a personal AI account, an organisation-managed education/business environment and a purpose-built application using an API is also hugely important. The underlying AI model might even be the same — while the data journey surrounding it can be very different.
Current UK Department for Education guidance specifically asks schools to consider how generative AI tools use personal data and recommends using approved tools, involving data-protection leads and avoiding personal data unless its use is necessary and properly protected.
This distinction is becoming increasingly important as education-specific AI environments emerge. For example, OpenAI states that information shared within ChatGPT for Teachers is not used to train its models by default and provides school/district administrative and privacy controls.
DEEP AI Audit
This is also exactly why we're building the DEEP AI Audit.
An AI policy tells you what should be happening.
An AI audit helps you understand what is actually happening.
Which AI tools are being used? By whom? For what purposes? What information is being shared? And where are the potential gaps in governance, knowledge and practice?
DEEP AI Audit: https://audit.deepeducationnetwork.com/
