AI access isn’t AI equity: what EX practitioners need to notice

A lot of workplace AI conversations still start with access.

Who has Copilot? Who has ChatGPT Enterprise? Which teams are using AI regularly? How many people have completed the training?

Those measures are useful. But from an employee experience perspective, they only tell us part of the story. Because access to AI does not necessarily mean access to the same opportunity. Two employees might have exactly the same tool available to them and have completely different experiences of what that tool makes possible.

One may work in a role where AI can remove hours of repetitive work, have a manager who encourages experimentation and have access to good quality information that AI can actually use.

Another may struggle to see how AI applies to their role, work in a team where the rules feel unclear and have very little time or opportunity to experiment.

Same organisation. Same tool. Very different experience. And that should matter to anyone working in EX.

The gap between individual and organisational readiness

Microsoft’s 2026 Work Trend Index gives us a useful way to think about this.

Its research with 20,000 AI users across 10 countries describes a “Transformation Paradox”: employees may be building their capability with AI faster than the organisations around them are adapting.

Only 19% of those surveyed fell into Microsoft’s “Frontier” group, where strong individual AI capability was matched by an organisational environment that enabled it.

Perhaps most interestingly for EX practitioners, Microsoft found that organisational factors such as culture, manager support and talent practices were associated with more than twice the reported AI impact of individual factors.

That shifts the conversation away from:

“Why aren’t some employees using AI?”

and towards:

“What is it about their experience that makes using AI easier or harder?”

You can explore the full Microsoft 2026 Work Trend Index here.

EX practitioners have seen this pattern before

This isn’t unique to AI.

Think about flexible working. An organisation can have one policy, while employees experience very different levels of flexibility depending on their role, manager and team.

Or learning and development. Giving everybody access to the same learning platform doesn’t mean everyone has the same time, confidence or encouragement to use it.

EX practitioners already understand that formal access and lived experience aren't the same thing.

So perhaps we need to apply the same thinking to AI.

Role is an obvious factor. Someone whose job involves writing, analysing, researching or synthesising information may quickly find dozens of useful applications for generative AI. Someone working predominantly with physical equipment or highly restricted information may find fewer.

That doesn't mean the second person is resistant to AI. It might simply mean the organisation has done less work to understand where AI could genuinely help them.

Managers can create another divide. One manager might actively share examples, encourage experimentation and create space for people to learn. Another might rarely mention AI or subtly signal that experimenting with it isn't "real work".

And then there's time.

Learning how to use AI well requires experimentation. Employees need opportunities to test things, make mistakes and work out what genuinely helps. If that learning is expected to happen on top of an already overloaded job, some people will inevitably have more opportunity than others.

The result is that two employees with identical access to technology can end up with very different AI experiences.

We may need better questions

This is where EX can add something important to the AI conversation.

Rather than simply measuring licences, logins or training completion, we can start looking at the experience underneath them.

Where is AI genuinely making work better? Which groups are benefiting most and least? Where do employees have permission and time to experiment? How much variation exists between teams? Are managers equipped to support their people? And are there roles where we need to redesign the work before AI can add meaningful value?

Perhaps most importantly, whose experience aren't we hearing about?

Because high adoption doesn't automatically mean high value. And low adoption doesn't automatically mean resistance. Sometimes the problem isn't the employee's willingness to use the technology. It's the environment we've asked them to use it in.

AI equity doesn't mean identical experiences

None of this means every employee should use AI in the same way.

Different roles will need different tools, support and levels of AI involvement.

For one employee, AI might mean sophisticated tools embedded directly into their workflow. For another, it could simply mean removing one frustrating administrative task.

Some people will need technical training. Others might benefit more from seeing examples relevant to their role. Some managers will need help creating safe environments for experimentation. And some teams may need better processes or information before AI becomes useful at all.

Equity isn't about making everyone's experience identical.

It's about making sure people have a fair opportunity to benefit.

And that's why EX has such an important role to play.

Perhaps the next evolution of the workplace AI conversation is to move beyond asking:

“How do we get more employees to adopt AI?”

and start asking:

“What would need to be true for more employees to benefit from it?”

That small shift changes the conversation considerably.

Because AI transformation isn't only about the technology we give people.

It's about the experience we design around it.