They’re not technical questions — the technical ones are easy to solve. They’re moral ones. And most of the industry is either answering them too slowly, or not at all. As leaders these are the ones we need to address the most.
The data problem no one wants to talk about
Start with consent. Platforms are training AI models on user-generated content — resumes, posts, conversations — data created for one purpose, now repurposed for another. This isn’t an edge case. It’s become standard practice. The precedent isn’t new either. We’ve already seen what happens when platforms collect and use data ahead of clear accountability. The difference now is scale and permanence. Once data is absorbed into a model, it doesn’t get “deleted” in any meaningful way.
When AI starts shaping identity
The more complex issue is what happens next. Generative AI isn’t just using our data — it’s beginning to shape how we think. These systems reflect human behaviour back to us. But not perfectly. They amplify patterns, biases, and emotional signals in ways that can subtly influence perception and decision-making. At scale, that matters.
When people, especially younger users, begin forming relationships with AI systems, the line between tool and influence starts to blur. This isn’t just product design. It’s behavioural impact. And yes I have seen this in real life, the way teenagers interact with AI. It is different. It can be scary.
The labour and creativity shift
The conversation around AI and jobs is often framed too simply. It’s not “jobs disappear” or “jobs remain.” It’s a redistribution. Certain categories of work are already being compressed — entry-level writing, design, and coding roles are contracting as AI handles first-draft output at a fraction of the cost. The people who built careers stepping into those roles are finding the ladder shorter than it used to be. At the same time, the qualities that define meaningful work — judgment, context, perspective — are becoming more important, not less. The disruption is real. And it won’t be evenly distributed.
The cost we don’t talk about
There’s also a physical layer to all of this. AI infrastructure consumes significant energy and resources. That doesn’t make AI inherently negative. But it does challenge the idea that it’s purely progressive. There’s a difference between applying AI deliberately to solve specific problems and deploying it broadly without constraint. One creates value. The other creates cost, often hidden.
What this actually comes down to
We are unlikely to regulate our way through this fast enough. The technology is moving faster than the frameworks around it. So the responsibility shifts. To builders. To leaders. To investors. Ethics cannot be treated as compliance.
The question isn’t just: “Is this legal?” It’s: “What does this do to people?”
We’re already seeing this play out in real deployments — strong technical capability, but unclear boundaries around data use, ownership, and accountability. That gap is where most of the risk now sits. Consent. Transparency. Developmental impact. Labour displacement. Environmental cost. These aren’t separate issues. They are different expressions of the same question:
Who does this technology actually serve? If we can’t answer that clearly, we’re not building the future. We’re just automating the present.
A closing note on how we work
This is why the engagement model at Immersive Realities starts with whether a technology should be built for a client, not just whether it can be. The commercial pressure is always to say yes. Someone in the room needs to be willing to say wait, or no. That’s not a marketing position. It’s an operating one — and it’s the reason we structure engagements the way we do.