Who Does Your AI Represent?

For the past two years, organisations have been learning how to work with AI in the workplace. Training programmes have been rolled out, prompt libraries created and new policies written. Leaders talk about adoption rates, productivity gains and responsible use. Employees are encouraged to experiment, while technology teams work to integrate AI into everyday workflows.

These are all important conversations. They reflect the reality that AI is becoming another workplace tool, much like email, spreadsheets or search engines before it. Yet during a recent conversation with Liz Villani, founder of #BeYourselfAtWork – Sounds Like Me® From iAM, I was asked a question that caught me completely off guard: “Do you see people talking about representation?”  

I paused. “No.”

It wasn’t because I’d never thought about authenticity or brand voice. Those ideas have become increasingly common as organisations experiment with generative AI. What surprised me was that I’d never considered representation as the missing piece of the conversation. Since then, I haven’t been able to stop thinking about it.

We’ve become very good at teaching people how to use AI. We’re spending far less time asking whether AI is actually representing the people using it.

The same technology, different organisations

One of the defining characteristics of this wave of AI is that the technology itself is becoming increasingly accessible. Whether an organisation chooses ChatGPT, Copilot, Gemini or Claude, the gap between those tools is often far smaller than the gap between the people using them. That changes where competitive advantage comes from.

If everyone has access to similar models, differentiation no longer comes simply from having AI. It comes from the judgement, experience and perspectives that people bring when they use it. The technology might be shared, but the thinking behind it shouldn’t be.

Ironically, as AI workplace communication becomes increasingly common, that is exactly what they risk losing if AI quietly becomes the default author of every email, proposal, presentation and customer response.

Not because the technology is poor, quite the opposite! AI is remarkably good at producing competent, polished work. But competence and distinctiveness are not the same thing. The more we rely on the same statistical patterns to communicate, the easier it becomes for different organisations to sound increasingly alike.

Years spent finding a voice

Most organisations invest enormous effort in defining who they are;

• Marketing teams develop tone of voice guidelines.
• Leaders work hard to build cultures that encourage different perspectives while aligning people around shared values.
• Recruitment focuses on bringing in people with different experiences because those differences lead to better decisions, stronger relationships and more original thinking.

Over time, those individual contributions become part of an organisation’s identity. Customers recognise it, employees feel it, competitors struggle to replicate it. Yet it’s surprisingly easy for that identity to become diluted;

• An email is rewritten.
• A proposal is polished.
• A customer response is generated in seconds.
• A LinkedIn post created with AI sounds just a little more generic than the last one.

None of those moments feels particularly significant on its own. Collectively, however, they raise an uncomfortable question; If AI is producing more of our communication, whose voice are customers, colleagues and partners actually hearing?

AI representation is more than tone of voice

At first glance, AI representation sounds like a branding exercise. But it is also closely connected to AI authenticity: whether the content AI produces genuinely reflects the people, perspectives and judgement behind it. It conjures images of writing style, preferred phrases or the difference between a formal and conversational tone. But that’s only a small part of it.

The most valuable people inside an organisation aren’t valuable because they know where to place a comma or which adjective best fits a sentence. They’re valuable because they exercise judgement. They understand context, recognise nuance and draw on years of experience to make decisions that another equally intelligent person might make differently. That is what clients trust, it’s what colleagues rely upon and it is often what separates one organisation from another.

Technology should help people apply that judgement more effectively. It shouldn’t quietly replace it with something statistically average.

A different conversation

Listening to Liz, I found myself wondering whether we’ve accidentally framed AI adoption too narrowly. We ask whether people are using AI, measure how frequently they’re using it and celebrate the productivity improvements that follow. Those questions matter, but they only tell us whether the technology is being adopted, they don’t tell us whether it reflects the judgement, experience and identity of the people using it.

Perhaps another question now deserves equal attention: not simply whether people can use AI, but whether AI genuinely represents them. That is where AI authenticity becomes important.

That feels like a very different challenge. It is less about introducing another piece of technology and more about preserving the qualities organisations have spent years trying to build.

If AI is going to become part of almost every interaction we have at work, AI representation may become just as important as adoption.

Final thought

Much of the AI conversation has focused on making technology more capable. Perhaps the next stage is making sure it never loses sight of the people it’s supposed to support. After all, organisations don’t build reputations through technology alone, they build them through the judgement, character and experience of the people behind it.

As AI-generated content becomes an increasingly visible part of workplace communication, AI authenticity and representation may become just as important as adoption.

If AI increasingly becomes the voice that customers, colleagues and partners interact with, then one question feels impossible to ignore. Who does your AI represent?