People Leaders think their job is to adopt AI. It isn’t. It’s to translate it.

Most organisations are sprinting into AI adoption based on the AI-hype narrative. That leads to a flurry of experimentation without purpose, testing without clear success metrics, and all of it happening in silos that limit communication of outcomes.

It looks productive. It feels progressive. In reality, it’s neither.

Because AI isn’t failing due to lack of effort. It’s failing because no one’s translating what’s actually happening.

Teams are building vertical capability without a horizontal foundation. Leaders are making decisions based on half-understood assumptions. And the language people use to describe AI work varies significantly from one department to the next.

People Leaders keep being told their job is to “scale AI adoption”. That’s the wrong job.

Your real job is translation.

You’re the only function with the vantage point across people, performance, workflow, capability, and culture.

You’re the only ones who can turn scattered experiments into something coherent.

If organisations want to move from single-player experimentation to multi-player capability, translation has to come first.

Here’s how.

1. Translate the Vertical pilots

Most companies have AI tools and pilots at different stages of test-&-learn across the business, creating the perception of progress without the coherence to back it up.

The instinct is to fix them. Don’t.

Just translate them.

Ask the questions nobody else is asking:

• Who’s using what?

• For which tasks?

• With what results?

• Based on which assumptions?

• And who actually knows this is happening?

This alone exposes the real state of AI inside your organisation. Spoiler: in every conversation I’ve had with People Leaders, it’s consistently more fragmented than anyone assumes.

Translation is not about controlling the pilots. It’s about understanding them well enough to make them useful.

2. Build the Horizontal literacy layer

Vertical work gets all the attention because it looks impressive. But without shared literacy, it’s meaningless.

The horizontal bar of the T-Model is your foundation:

• One shared language

• One set of definitions

• One organisational understanding of what AI is and isn’t

This isn’t a training programme. It’s not a vendor playbook. It’s the baseline the entire company stands on.

Without it, every pilot becomes a separate dialect no one else can read.

3. Create a shared organisational language

Most AI problems inside companies aren’t technical. They’re linguistic.

Teams use the same words to describe completely different things.

Leaders make decisions based on imaginary outcomes.

AI “wins” sound plausible in stand-ups and fall apart when examined.

People Leaders need to define the language:

• What counts as a use case

• What counts as value

• What counts as fluency

• What counts as risk

• What counts as adoption

• What counts as noise

Shared language is the glue between vertical depth and horizontal inclusion. Without it, organisations default to chaos disguised as progress.

4. Map the fragmentation before scaling anything

Give yourself two hours and a whiteboard. Map everything:

• Tools being used

• Tasks being automated

• Who’s doing what

• Where the duplication is

• Where the blindspots sit

• Where assumptions are driving decisions instead of evidence

This is your translation layer in practice.

Most companies discover they’ve accidentally built five versions of the same workflow, each in a different team, none talking to each other.

You can’t scale what you haven’t mapped. And you can’t lead if you can’t see.

5. Re-sequence without losing momentum

Once you’ve translated what’s already happening without interfering, your next move is to re-sequence the order of learning so the organisation can progress without losing momentum.

People Leaders often get told: “Fix this.” “Standardise that.” “Slow everything down until we understand it.”

Don’t stop everything. Re-sequence.

Stopping vertical experiments reduces confidence and impacts momentum.

Your job is to support what’s already happening while re-sequencing the order of learning beneath it.

• Keep the pilots running.

• Build the horizontal layer at the same time.

• Translate the work into something the wider business can understand.

• Shift from individual wins to shared capability.

This is how organisations move from hype to habit.

Single-player mode gets early wins. Multi-player mode compounds them.

And People Leaders create that shift.

Your job now is translation

Your organisation doesn’t need another AI programme. It needs someone to make sense of the one it already has.

This is the People Leader agenda now. You define the order of learning before the organisation defines it for you.

That foundation is yours to build.

If People Leaders don’t create shared language, coherence, and sequence now, someone else will. And they won’t do it well.

Your job isn’t to adopt AI. Your job is to translate it.

Start now.