The AI for Good Global Summit in Geneva is the United Nations’ premier platform for ensuring artificial intelligence benefits humanity and has become one of the key international gatherings where governments, industry leaders, researchers, and civil‑society organizations converge to shape how artificial intelligence is deployed for public benefit. Framed around practical applications, the summit pushes forward real‑world collaborations that can influence how security, safety, and resilience are understood worldwide. With discussions ranging from emerging regulatory frameworks to the operational realities of AI in frontline environments, the conversations are shaping how private security organizations will operate in an increasingly AI‑enabled landscape.
Question: After having the opportunity to observe those conversations firsthand, what was your big takeaway?
Stefan Huber, Executive Dir., International Security Ligue: I came away with a significant conviction: that AI itself is no longer the biggest challenge. Trust is.
Everyone talks about larger models, faster chips, and new applications. Yet AI has no purpose of its own. So, before asking "Which AI should we use?", perhaps every organisation should first ask: "What outcome are we trying to achieve?"
Q: How do you see that applying to the security domain?
SH: One topic repeatedly surfaced for me: convergence. Too many organisations still see it as connecting cameras, access control, cyber or AI. That is integration. Real convergence connects people, AI, physical security, cyber, governance and trusted information into one operating model. Technology evolves exponentially. Governance does not. Bridging that gap is becoming a leadership challenge.
One discussion that particularly stayed with me was "AI Impact at Scale," hosted by the Permanent Mission of India to the United Nations, with an excellent fireside discussion moderated by Rudra Chaudhuri. The examples demonstrated genuine impact—but none are yet truly sustainable at scale.
Q: What is holding us back?
SH: Financing, certainly. And possibly because a lever for getting financed — a critical accelerator hardly any would think of — is still largely missing: Procurement.
Q: What do you mean?
SH: Governments and policymakers don’t just regulate markets. Through the way they structure and carry out public procurement, they influence how entire industries develop.
This is particularly relevant in India right now, as it modernises its labour legislation. The private security sector is no longer an “unskilled” field, a classification that has resulted in misdirected subsidies. In fact, it is evolving into an increasingly professional, technology‑enabled part of national resilience.
And AI can accelerate that transformation — but only if procurement evolves as well. Procurement data should become structured, machine-readable, and based on measurable quality, not simply the lowest price. AI can only reward quality if quality can actually be measured.
Q: You mentioned the increasing role trustworthy AI in the technology landscape, what got you thinking about that?
SH: One of the discussions really challenged my thinking: the Trust Breakfast – AI, Information Warfare and Multilateral Trust, hosted by Axel Mazolo, GAIGI / AI Trust House, together with a really interesting mix of participants. It focused on information trust.
Information itself has become a battlefield. The objective is often no longer aimed at stealing information — it is about undermining confidence in governments, media, institutions and, ultimately, each other.
Information itself has become a battlefield.
And led me to the conclusion that the next strategic competition is no longer for data. It is for trust. Trustworthy AI is therefore less about selecting the most powerful model and more about transparent governance, ethics, accountability, and human oversight.
Q: I guess that in a world where data is everywhere, the information that truly matters is what manages to rise above the noise.
SH: One example has stayed with me. We still remember the Wright brothers, but not because they were the first to achieve sustained flight, but because their famous 1903 flight was also successfully captured on film and is easy to find and freely available.
The AI era follows a similar logic. As AI increasingly replaces search with generated answers (Goolge bar), organisations that produce structured, machine-readable, and trusted information will increasingly shape the answers. The visual polish of a website matters less than the credibility of the information behind it. In this environment, being a consistently authoritative source that AI can recognize becomes a genuine strategic advantage.
Q: What is your final takeaway?
SH: I think it’s that everyone is racing to build better AI, but perhaps we should spend more time building better trust. Because AI will not define the future. The confidence people place in it will.
