On paper, this marks real progress — and it is.

For years, young people have argued that artificial intelligence is not a distant policy concern but something already woven into our classrooms, friendships, work, and access to information. Institutions, it seems, are beginning to catch up.

Still, I find myself returning to a different question — one that sits slightly beyond the conference halls and the communiqués: what outcomes are we actually trying to create?

Much of today’s AI governance debate is necessarily anchored in the present. We are concerned with misinformation, algorithmic bias, exploitative data practices, and the growing number of digital harms affecting children and young people. Those concerns are urgent. But governance built exclusively around current risks has a short shelf life.

Twenty years ago, few policymakers could have anticipated the ways social media would reshape political participation, childhood, public trust, or even the idea of community itself. The internet we inherited was not the internet that had been imagined. There is little reason to believe that artificial intelligence will unfold more predictably.

As young people, we occupy an unusual position in this conversation. We are often described as “the future,” yet we are already living with the consequences of decisions being made today. The systems being designed now — the safeguards, institutions, norms, and accountability mechanisms — will outlast the news cycle that produced them.

That changes the question.

The outcome I want is not simply safer AI, though safety matters enormously. Nor is it a world in which young people are invited into governance processes only after the most consequential decisions have already been made.

I want institutions capable of adapting to technologies that do not yet exist.

I want youth participation to move beyond consultation and toward genuine co-design — not because representation is inherently virtuous, but because governance systems are stronger when they incorporate the experiences of those who will live with their consequences the longest.

I want AI literacy to mean more than technical fluency. Knowing how to use a model is one thing; understanding how systems are governed, whose interests they serve, and where accountability lies is another entirely.

And I want conversations about AI governance to become less geographically concentrated. Some of the most important questions about technology, education, labour, and human rights are being asked in Nairobi, Bissau, São Paulo, and Dhaka, often by communities that encounter AI not as developers or investors, but as citizens navigating institutions that were never designed with them in mind.

The recent developments in Geneva suggest that international governance is beginning to take youth perspectives more seriously. That is welcome. But inclusion alone is not an endpoint.

The real test will come later: when the technologies change, when new risks emerge, and when today’s frameworks are asked to govern circumstances no one fully anticipated.

Because, ultimately, the challenge before my generation is not whether artificial intelligence will shape the future. It already does.

The question is whether the governance systems being designed right now adequately account for their own limits — or whether we are, once again, building for the world we can imagine rather than the one that is actually coming.