AI transparency: a gold star for schools.

How schools are ahead on AI transparency

There’s a framework being used in education that could help people in the corporate world. 

The Artificial Intelligence Transparency Scale (AITS), from Sheffield Hallam University.

It’s a way of describing the role AI played in a piece of work. It’s a 1-5 scale that ranges from no AI involvement, through to using it for all elements of a task. 

Students say how they’ve used AI, and assessors use that knowledge when reviewing the work.  

Why it’s needed in education

As AI-generated writing becomes more fluent, it’s getting harder to see where thinking is happening and where it isn't. 

It can mask: 

• gaps in understanding 

• borrowed reasoning 

• a lack of decision-making.

Not helpful for teachers trying to assess their students’ work. 

And it’s a problem for students too. Research suggests that heavy reliance on AI can erode critical thinking skills over time. So educators need to encourage AI use in a way that keeps the cogs turning.

That’s the same deal in the corporate world

Writing that sounds fluent and looks polished, can hide weak thinking. 

Like this…

This initiative represents a strategic step forward in how we support our teams to work more effectively. By leveraging new tools and ways of working, we’re creating space for better collaboration, clearer outcomes and more consistent decision-making across the organisation.

Looks okay, sounds like something a business would say. But then you start to see what’s missing. You’re left wondering:

What initiative?

Which tools?

What’s changing exactly?

Just like schools and universities, businesses

need to encourage people to stay in the loop and do the thinking.  

A shared language around AI use

A recent Forbes article points out that AI usage policies will become increasingly important. No surprise, when you’ve got surveys like this saying 57% of people admit to inputting sensitive company data to publicly available AI tools. 

So the focus of AI usage policies will be on permission and risk. What’ll likely be missing is a shared vocabulary.

Something that lets people say…

“We used AI for shaping — it helped us explore strategy directions."

“We used AI for enhancing — it co-authored parts of the report.”

That kind of clarity helps in a few practical ways.

Managers don’t have to guess where the thinking came from. Reviewers know where to focus their attention. And authors who decide upfront how they intend to use AI will make better choices as they go.

Having a clear way to talk about the use of AI makes the thinking easier to see.