Redefining Digital Inclusion for the AI Economy

Written By Vic MacLennan

For the past decade, the national conversation around digital inclusion in Aotearoa has rested on a straightforward premise: if we can bridge the basic access, skills and trust gaps, we can build a level playing field.

In simple terms, we’ve spent ten years treating digital equity like a pipeline problem—get a laptop, get a connection, teach some fundamental skills, job done. That work was vital, and the community of practice building those bridges has laid a foundation we couldn't live without.

But Artificial Intelligence just blew that framework apart.

As AI rapidly reshapes global economies, workplaces, and public services, we are witnessing a fundamental shift in what "inclusion" actually means. This is no longer just about who can log on to check emails, submit a government form, or help their kids with homework.

The question is shifting from access to agency: Who has the capability, capital, and power to make technology work for them—and who is simply subject to its decisions?

If we are serious about addressing this new landscape, we have to start with an uncomfortable truth: achieving equity requires an unequal allocation of resources to secure fair outcomes. Simply handing everyone the same baseline tool kit will not prevent systematically underserved communities from falling further behind in an AI-driven economy.

The UN recently warned that the concentration of AI compute infrastructure in a handful of wealthy regions threatens to deepen global inequalities, creating a sharp divide between those who own the systems and those who absorb their systemic costs. Meanwhile, the OECD cautions that generative AI risks widening regional and socioeconomic gaps if lower-income communities are left with outdated tools.

As far as I can tell, while the broader evolution of artificial intelligence will impact society in countless ways, we can already see this new divide opening up across three critical fronts in today’s generative AI landscape:

1. Infrastructure & Compute: Beyond the Five-Year-Old Smartphone

A basic smartphone on a capped prepay data plan was enough to participate in the digital world of today. It won't be enough for the world we are entering.

AI runs on enterprise-grade compute, low-latency broadband, and heavy industrial infrastructure whose energy and water footprints often fall disproportionately on regional environments. Without deliberate investment in modern tech for under-resourced households and community hubs, this hardware divide will turn into a permanent wall. The issue is no longer just bridging the connectivity gap; it is ensuring equitable access to raw compute power while confronting the massive carbon, water, and energy footprints these systems extract from our local environments.

2. Work & Opportunity: The Missing Rungs on the Career Ladder

For generations, entry-level roles—administration, customer service, junior analysis, service desk support—have served as vital launchpads into professional careers. They are how young people and underrepresented demographics build networks, gain confidence, and move upwards.

Generative AI is beginning to automate these exact early-career tasks. I’m not suggesting these roles will vanish overnight, but if the first rungs of the career ladder are erased, how do people get their start? There is a real risk that while the efficiency gains of AI accrue to senior, highly-capitalised professionals, those entering the workforce will find the ladder pulled up ahead of them.

3. Governance & Power: Who Designs the Future?

This is where equity meets agency—and where digital inclusion becomes about power.

Most commercial AI systems learn from massive overseas datasets that reflect someone else's biases. If we roll them out here without genuine local oversight, we risk taking old forms of discrimination and baking them right into our automated future.

When automated tools are brought in to help screen job applicants, assess credit ratings, or streamline public services, the rules built into those algorithms have immediate real-world impacts. Many of these systems rely on datasets gathered offshore that carry their own built-in bias and assumptions. Without local oversight and thoughtful governance, it's all too easy to inadvertently import those old biases and weave them straight into our everyday operations. 

Real governance can’t be something that happens after the fact; it needs to be about making sure our communities have a seat at the table and a genuine voice in how these tools are chosen, built, and held accountable. Otherwise the tech companies will hold all the power. (Lecture over).

New Goals

Over the last decade, we rightly focused on getting people connected. But if we continue to measure success by yesterdays metrics, we risk solving 'Digital Equity 1.0' while leaving our communities entirely defenseless against the divides of 'Digital Equity 2.0 - AI.

For those of us in the digital inclusion space, our missions need to evolve. We still need to ensure communities have access, skills and trust but our goals can’t end there; we need to ensure our communities have the agency, resources, and power to shape and benefit from AI. Yes this is a much bigger challenge—but if we want an Aotearoa where everyone can thrive in an AI-enabled economy, it’s the work we need to take on. I have barely scratched the surface here so expect more on the AI divide very soon. Vic

Reference articles:

Vic MacLennan
Next

The government marked its own homework on grocery labels.