Which AI tools are your staff using?

[dhicks]

25 minutes ago, DavR said:

NotebookLM looks interesting too, I like the idea of creating your own limited knowledge base, rather than using the whole LLM data. Time will tell if that's something I could sell to others.

 

We currently have inspectors in, doing a pre-inspection run-through. Everyone is being reminded to look at our policy documents, which of course are impractically large to simply read through. A NotebookLM of all the policy documents, shared around staff and available for them to ask questions of as needed, might be a really nice use-case. The same, possibly, with our technical support documentation. We have a bunch of (hopefully well-written) documents describing how to go about various tasks / processes. Piling all those into a NotebookLM and, again, letting the staff ask questions potentially gives us a technical support assistant.

 

A problem with NotebookLM seems to be that it won't scan for changes to a folder of documents - you have to manually remove the old document and add the new version. This looks like it should be scriptable, though, so we can maybe have something scan for changes daily and update Notebooks as needed.

[flipthebit]

16 minutes ago, dhicks said:

 

We currently have inspectors in, doing a pre-inspection run-through. Everyone is being reminded to look at our policy documents, which of course are impractically large to simply read through. A NotebookLM of all the policy documents, shared around staff and available for them to ask questions of as needed, might be a really nice use-case. The same, possibly, with our technical support documentation. We have a bunch of (hopefully well-written) documents describing how to go about various tasks / processes. Piling all those into a NotebookLM and, again, letting the staff ask questions potentially gives us a technical support assistant.

 

A problem with NotebookLM seems to be that it won't scan for changes to a folder of documents - you have to manually remove the old document and add the new version. This looks like it should be scriptable, though, so we can maybe have something scan for changes daily and update Notebooks as needed.

How well/accurately were questions answered from the store of policies?

[dhicks]

10 minutes ago, flipthebit said:

How well/accurately were questions answered from the store of policies?

 

From a quick test, it seemed accurate, although this is the kind of thing that really needs testing in practice - it answered my bunch of test questions okay, but how it handles the sort of thing users throw at it I'm not sure. The first problem we encountered was collating all the appropriate files to put into the Notebook - some are in one folder, some in another, this is the kind of thing we'll need to take a bit of time to have a proper sort out with. I was quite impressed by the "Further questions" feature - at the bottom of a generated answer, you generally get two or three "Have you also thought of asking...?" questions, which was actually quite useful for the policies Notebook - you can browse though linked questions better than trying to read a 100-page policy document.

[TwistedHelixis]

Primary school are using Google Vids, which is powered by AI to create videos from their Google docs.

 

Gemini is used by the staff, but all the main Ai systems including Gemini are blocked for the pupils.

 

Didn't really like Copilot and ChatGPT is a no

[steve]

I think my org now have some reasonably comprehensive guidance for all our users: Generative AI | Guidance, policies and information about using Generative AI for staff and students. This policy page might help in creation of your own: Guidance to Staff on the use of Artificial Intelligence | Governance and Compliance

 

We also have mandatory training now too on AI use - basically co-pilot for anything sensitive is a must.

 

I'm currently working on a project to deliver most other AI models securely via Azure AI Foundry to try prevent (limit) data exfiltration. The uptake stats and cost of this will hopefully come from a good pilot scheme.

 

But its a tricky one no matter the org size and funding available.

[msi_school]

On 14/01/2026 at 13:41, flipthebit said:

How well/accurately were questions answered from the store of policies?

My favourite feature of Notebook LM is that it provides links to the original data in it's answers so you can double check. 

[Gheedn]

On 13/01/2026 at 14:20, DavR said:

We're just in the process of creating our policies on AI use in school, and that will include which AI tools are going to be our recommended tools, and which are not. From a poll we ran just before Christmas, we found that at the moment, most people who are using AI are using ChatGPT. 

 

We have the licensing for entry level Microsoft Copilot and Google Gemini, but I'm just wondering whether it's worth our time getting these tools set up and tested, when so far most people have gone to ChatGPT. I'm sure we've all been there before and put a lot of time and effort into a technically better product, only for users to go with the "buzz word" product of the day!

 

As a third option, our Headteacher is also keen on Teach Mate, a school specific AI tool. This would be an extra cost, but, it has a lot of school tasks built in to it. If we bought into that, then we'd very much be pushing our staff to use it.

 

What AI tools are your staff using? Have you allowed them to use popular tools of their choice such as ChatGPT, or, have you sought to lead behaviour towards specific school approved tools, and if so which ones? I tried the perchance ai video generator   and was impressed by how fast it turns prompts into engaging clips. Great for quick social content or lesson visuals without editing skills. Not perfect for complex storytelling yet, but superb for brainstorming and inspiration. Easy to use and fun.

We faced a similar debate. Most staff naturally gravitate to ChatGPT because it’s familiar and flexible, so banning it often just drives use underground. Our approach was to allow popular tools (like ChatGPT) with clear guidance, while recommending approved options such as Copilot or Gemini for data-sensitive work. School-specific tools can add value if they genuinely save time, but forced adoption rarely sticks. We also see creative experimentation with tools like the Perchance AI video generator, which can support media and lesson ideas if boundaries are clear. Clear policy > forcing one tool.

[MYK-IT]

Ai, and use of Ai within schools is certainly far from being straight forward and is creating a lot of work for many and policies too! From the teaching staff who want to teach, use and embrace Ai for T&L, and for those doing their due diligence (e.g. DPL/DPO - DPIA & Data Protection, Safeguarding Lead - safeguarding, IT/Digital Lead for filtering, reporting and monitoring.). Some of which I don't think is achievable (yet), or certainly without implementing additional filtering/services/products (or funding!) etc.

 

For example,.

• Ai prompts need to be logged - However, M365 Copilot Chat does not allow M365 Admin to see what prompts a user has entered etc.
• Ai Generated Images - where do you start on monitoring if such Ai generated content is appropriate, or not?
• Need to protect the student's IP (e.g. school work, art etc)
• PII Data Protection - Although we state to use approved Ai only (e.g. M365 Copilot Chat) the user remains the weakest link to potentially entering PII data into an Ai (model) that is not approved / enterprise protected. Although you could argue that this would be covered by an existing Data Protection Policy

 

We have a provisional Ai Policy, but we already know it needs much more consultation and amendment - As use of Ai is dynamically changing, as is the DfE's guidance! Which was updated only a couple of days ago to include reference to the Online Safety Act. Generative AI: product safety standards - GOV.UK

 

In relation to OFSTED, How Ofsted Looks at AI During Inspection | Derbyshire Services for Schools is a useful reference.

 

In relation to ChatGPT, we have been advised it will not pass a DPIA, hence our policy reflects accordingly.

 

Edited January 21 by MYK-IT

[Ditto]

Fundamentally, Copilot is ChatGPT under the hood. I'm interested what it is about ChatGPT that failed your DPIA process.

[MYK-IT]

12 minutes ago, Ditto said:

Fundamentally, Copilot is ChatGPT under the hood. I'm interested what it is about ChatGPT that failed your DPIA process.

 

Was advised that ChatGPT* has no formal data processing agreement, age verification and limited control as to where and how data is stored. This could lead to a potential data breach or unauthorised use/processing.

 

(*Free and Plus versions - I know that the likes of ChatGPT Business and ChatGPT Edu exist and have DPA etc, but they are at additional cost and EDU version is for university-aged students, on upwards. Hence, our use of M365 Copilot Chat as already licensed, and can implement age-controls etc).

[synaesthesia]

As above - you have no option to control storage or processing of data with ChatGPT, nor can you tell it not to use your data to "improve it's algorithms". You can "control" what data is used by policy, but that relies on humans doing what they're told and not uploading documents full of PII to it.

Once it's there, it's effectively a data breach.

[DavR]

6 hours ago, Gheedn said:

We faced a similar debate. Most staff naturally gravitate to ChatGPT because it’s familiar and flexible, so banning it often just drives use underground. Our approach was to allow popular tools (like ChatGPT) with clear guidance, while recommending approved options such as Copilot or Gemini for data-sensitive work. School-specific tools can add value if they genuinely save time, but forced adoption rarely sticks.

 

This is probably the approach they we'll be taking I think - have Copilot and Gemini configured and available for anything involving personal data, but allow ChatGPT with strict guidance. 

In fairness, from a survey we did, most staff do seem to be quite switched on to the data protection risks, but that's only what people will admit to on a survey.