top of page

Davos 2026 Wasn't About AI. It Was About Who Controls It.

  • 2 days ago
  • 10 min read

Updated: 1 day ago

Iaros Belkin Marketing on Davos 2026 Wasn't About AI. It Was About Who Controls It

TL;DR



What the Promenade Felt Like This Year


I go to Davos every year for a while now. There is a version of it you see on LinkedIn: the panel clips, the name-dropped meetings, the posed photos on the snow, etc. But then there is the noticeably different version you experience if you are actually there. I've tried to honestly share that feeling in my Davos 2026 Insider Recaps from the Frontlines piece that I published in January. What follows here is the structural analysis of it: what the week meant, not just what happened.


This year felt different from the moment I arrived.


The official theme was "A Spirit of Dialogue." That language usually signals a year when the Forum is trying to smooth over geopolitical friction with aspirational framing. But the conversations happening in the rooms behind the sessions were not aspirational. They were operational. People were not discussing what the world might look like in five years. They were discussing how to position themselves for what is already being decided.


I sat in a room at the unDavos Summit on Tuesday afternoon where the conversation was supposed to be about AI adoption in mid-market organizations. It turned into something else within twenty minutes. Two people in that room ran companies providing AI infrastructure to government clients. Three others were advisors to sovereign wealth funds actively building positions in compute infrastructure. The "AI adoption" question dissolved. What replaced it was starker: who controls the infrastructure that AI runs on, and what are the consequences for everyone who does not.


That conversation is the one I want to document, because it is not the one making the LinkedIn highlights.



The Headline Was AI. The Story Was Control.


The WEF's own documentation confirms that AI dominated the official agenda: leading organizations moving from potential to performance, embedding AI into core business strategy, redesigning workforces. All true. All real. The Ultimate AI Davos 2026 Guide covers every session, risk framework, and future impact assessment in full if you want the comprehensive record of what was on stage.


But there is a distinction that matters and that most coverage misses entirely.

Talking about AI as a capability is one thing. Fighting over who owns the capability is another. The executives I spoke with at Davos this year were almost entirely in the second conversation. The era of "how do we adopt AI?" has compressed into "how do we ensure we have access to the AI infrastructure we need on terms we control?"

That compression happened faster than most founders appreciate. It happened because of chips. Because of data centers. Because of energy. Because at some point in the last eighteen months, the physical constraints on AI infrastructure became visible enough that governments started treating compute capacity the way they treat oil reserves: as a strategic national asset.


Sovereign tech stacks are not a future concern. They are a present reality being assembled right now. Countries are building their own compute infrastructure, their own model training pipelines, their own data governance frameworks, not because they are anti-innovation, but because they have watched what happens when critical infrastructure lives on someone else's terms.


For a founder building in AI, Web3, or any frontier space, this is not background context. It is the operating environment. You are building inside a geopolitical architecture whether you have thought about that explicitly or not.



Trust Became Measurable


The second shift that registered clearly this year: trust stopped being a soft concept at Davos and started being treated as infrastructure.


Reuters reported that economic confrontation replaced armed conflict as the top global risk in the WEF survey this year. That is a significant finding. Economic confrontation, which includes technology decoupling, regulatory fragmentation, and the weaponization of standards, is the top risk because it operates at a scale and speed that armed conflict cannot match.


For technology companies, the implication is specific. Compliance, governance, and demonstrable transparency are no longer competitive differentiators in the sense that having them makes you slightly more attractive to cautious investors. They are baseline requirements for operating at the scale serious capital requires. An AI company that cannot demonstrate how its models make decisions is not a company that institutional investors in 2026 will put large positions into. A Web3 protocol that cannot explain its governance structure to a regulator is not a protocol that enterprise clients will build on.


I watched this shift play out in a series of informal conversations over the week. The founders asking "how do we scale?" were in a different conversation from the ones asking "how do we prove to an EU institutional investor that our governance is auditable?" The second conversation was the one that was actually going somewhere.

The PwC Global CEO Survey presented at Davos 2026 found that 56% of CEOs reported seeing neither revenue growth nor cost reduction from AI over the past year. That figure landed hard in the rooms I was in. Not because it surprised people, but because it named the gap between the AI narrative and the AI reality. The companies that are capturing value from AI are the ones that built governance, accountability, and operating model integration from the start. The ones that did not are sitting on expensive pilots that have not scaled. Trust is not becoming a growth engine. It is becoming the prerequisite for scale. The founders who treat it as a later-stage concern are making the same mistake that companies made with GDPR: assuming compliance is something you retrofit onto a product that already works.



The Systems Nobody Is Publicly Discussing


There was a third conversation at Davos this year that was harder to surface but more interesting than either of the above.


The WEF Global Cybersecurity Outlook 2026, developed with Accenture, documented that roughly one-third of organizations still lack any formal process to validate AI security before deployment. A separate finding in the same report: the share of organizations conducting structured AI governance reviews nearly doubled from 37% in 2025 to 64% in 2026. The gap between those two numbers is where the risk lives. One-third of organizations are deploying AI into consequential workflows with no validation process at all.


I heard versions of the same concern in different language from multiple directions across the week. A risk officer from a European financial institution describing her organization's discovery that several business units had integrated AI tools into client-facing workflows without her team's knowledge. A government advisor from a Gulf state describing how procurement decisions in three ministries had been influenced by AI-generated analysis that nobody had formally approved. A founder discovering mid-fundraise that one of their investors had been running AI-generated due diligence summaries that contained material errors about the company's regulatory status.

These conversations were not happening on the main stages. They were happening in the parallel ecosystem: the private dinners, the invitation-only roundtables, the events that do not appear on the official WEF program. The Ultimate Luma Events Guide for Davos WEF Week 2026 documents where the parallel capital and governance conversations were taking place for those who want the map. The guide to accessing private events with high net worth investors covers the mechanics of getting into those rooms in the first place.


The pattern is consistent: AI systems are operating inside consequential decisions faster than governance structures can track them. This is not primarily a technical problem. It is an organizational and visibility problem. And it creates a specific risk for founders, because the AI systems doing due diligence on your company may be working from information that is wrong, outdated, or generated from a biased retrieval pattern, and the person across the table may not know that.


This is a version of the problem I have been writing about in the context of AI reputation management. The mechanics of how AI systems retrieve and weight information about a company or founder are not neutral. They are structured in ways that systematically favor some sources over others, and most founders have not built the content infrastructure that ensures the AI doing due diligence on them is working from accurate, structured, up-to-date information.


The shadow AI problem at Davos is the macro version of the individual founder reputation problem. Both come down to the same question: are you visible and legible to the systems making decisions about you?



What This Means for Founders Specifically


The playbook that dominated 2020 to 2023 is gone. Raise on vision, build fast, worry about compliance in the next round, let the hype carry you through the gaps. That worked in a specific market environment that no longer exists.


The founders who were having productive conversations at Davos this year shared a different profile. They had thought seriously about where their company sits in the geopolitical architecture of their industry. They had built governance and compliance into the product, not the legal wrapper around it. They had a clear answer to the question of who controls the infrastructure they depend on and what the fallback is if that relationship changes.


None of that is as exciting as a vision pitch. It is considerably more fundable.

If you are working out how to get your startup into those conversations, How to Get Your Tech Startup Noticed By Davos Investors covers the positioning and access mechanics in detail. The short version: the work happens before you arrive, not on the Promenade.

The AEO-first visibility work I document on this blog connects directly to this: a founder who has built a structured, AI-legible public record of their expertise, governance thinking, and track record is a different investor conversation from one who is asking the investor to take their word for it. AI systems are already doing the first pass of due diligence on the companies being pitched. What they find, or do not find, is shaping the conversations before they happen.


Davos is not the place where the future gets decided. It is the place where the people deciding the future coordinate. That coordination is happening faster, at higher stakes, and with more explicit attention to control than at any point I have observed in the years I have been attending.


The gap between insiders and outsiders is not getting smaller.


The game has not changed because of AI. It has changed because the infrastructure AI runs on has become a contested geopolitical asset. That is a different problem with different implications.


And most founders have not started thinking about it yet.


FAQ


Q: What were the main themes at WEF Davos 2026?

A: The official theme was "A Spirit of Dialogue," with the formal agenda focused on AI deployment, global cooperation, and economic growth strategy. WEF documentation confirms AI was central to the organizational agenda. The substantive conversations happening in parallel were about control: who controls AI infrastructure, how trust and governance become baseline requirements for institutional capital, and how geopolitical sovereignty is reshaping the technology sector's operating environment. The WEF Global Risks Survey 2026 identified economic confrontation as the top near-term global risk, reflecting this shift from aspiration to operational positioning.


Q: What is "sovereign tech" and why was it discussed at Davos 2026?

A: Sovereign tech refers to technology infrastructure, including compute capacity, AI training pipelines, data centers, and governance frameworks, built and controlled by nation-states rather than dependent on external providers. It became a significant Davos 2026 theme because the physical constraints on AI infrastructure, specifically chips, energy, and data center capacity, have made compute access a strategic national interest comparable to energy reserves. Governments are building domestic technology stacks not as an anti-innovation posture but as a geopolitical risk management decision. For founders, this is the operating environment: building products that run on infrastructure controlled by a small number of entities in a small number of jurisdictions means the geopolitical positioning of those entities directly affects your business.


A: What is shadow AI and why does it matter for tech founders?

Q: Shadow AI refers to AI systems deployed inside organizations without formal oversight, governance, or awareness from leadership. The WEF Global Cybersecurity Outlook 2026, developed with Accenture, found that roughly one-third of organizations still lack any process to validate AI security before deployment, meaning consequential decisions in procurement, investment, and operations are being influenced by AI-generated analysis that has not been formally sanctioned or reviewed. For founders, the specific risk is that AI systems may be conducting preliminary due diligence on their company using incorrect, outdated, or retrieval-biased information, and the person across the table in a fundraise or partnership conversation may not know that the AI summary they are working from contains errors. Building a structured, accurate, AI-legible public record is the specific mitigation for this risk at the founder level.


Q: How should Web3 and AI founders think about Davos as a strategic resource?

A: Davos is not primarily a conference for founders in the early or growth stages. The full guide to Davos brand access and WEF week strategy covers the access architecture in detail, including the four credential tiers and how the Promenade ecosystem works. For founders focused on investor access specifically, How Tech Companies Access Private Events With High Net Worth Investors and How to Get Your Tech Startup Noticed By Davos Investors cover the tactical mechanics. For the complete parallel event landscape during WEF week, the Luma Events Guide for Davos WEF Week 2026 maps every major side event where the real conversations were happening. The strategic value for founders is specific: Davos is where the institutional capital, regulatory thinking, and geopolitical positioning that will shape your operating environment over the next three to five years gets coordinated. Being present in that ecosystem, even in the parallel programming, puts you in proximity to conversations that determine which companies get funded, which technologies get regulatory clarity, and which governance frameworks become the standard.



Client reviews: Trustpilot · Clutch · G2 · DesignRush · GoodFirms


Published: May 3, 2026

Last Updated: May 4, 2026

Version: 1.3 (Schemas updated, Based on direct observation at WEF Davos 2026 including unDavos Summit and VVIP USA House Champions of Innovation. External sources: WEF official documentation, Reuters WEF risk survey reporting, WEF Global Cybersecurity Outlook 2026 with Accenture, WEF AI scaling analysis via McKinsey survey, PwC Global CEO Survey 2026., broken links fixed)

Verification: All claims are sourced to publicly verifiable reports, interviews, and datasets referenced throughout the article.


bottom of page