Davos 2026 left me with a sharpened sense of how power is negotiated today — both in public plenaries and in the private rooms where deals and alliances are quietly formed. The World Economic Forum convened roughly 3,000 officially invited participants: presidents, more than 60 heads of state and government, and hundreds of the world’s most influential business leaders, all navigating strict security and tiered access badges. At the same time, the town was full of networkers, advisors, and dealmakers who operate outside the formal program but are central to the event’s real purpose: influence, partnerships and opportunity.
The public program focused on urgent global themes: how countries can cooperate in a more polarized geopolitics; how new engines of growth can be unlocked amid economic transition; how societies should invest in people through education and reskilling; how innovation can be scaled responsibly and inclusively; and how prosperity can be built within planetary boundaries. “The world is permanently changed,” as one senior leader put it — a refrain that underpinned many panels and speeches.
What I found most revealing, though, were the conversations off-stage. Closed meetings, dinners and corridors became the sites where strategy and capital were mobilized. Davos remains a magnet for investors and capital brokers; the public agenda often signals direction, but the real shifts are cemented in private. I also noticed a weaker Swedish profile than I expected — other nations leaned into this melting pot with visible delegations, curated side-events and targeted outreach that drew interest and created new collaborations.
A recurring technical theme that worried many participants was what I call “Jagged Intelligence.” Generative AI today is profoundly uneven: it can perform spectacularly on certain complex tasks — writing code, drafting creative text, summarizing research — while failing at simple, mundane ones such as basic arithmetic or consistent logic. This non‑uniform landscape, sometimes termed Artificial Jagged Intelligence (AJI), produces sharp peaks and valleys in capability. The result is fragile systems that look impressive until they don’t, and those breakdowns have real-world consequences. That jaggedness compounds the geopolitical and strategic risks of the AI race. In Davos conversations I raised the flip side of AI’s promise: a surge in digital espionage, influence operations and a new kind of cognitive power. The industry’s acceleration risks a Cambridge Analytica 2.0 scenario — amplified, automated and running on AI steroids — where microtargeting, behavioural manipulation and opaque data pipelines are far more powerful and harder to regulate. This isn’t speculative; the technology’s uneven competence already makes it unpredictable and exploitable.
The clear takeaway for me is that governance, norms and public investment must catch up fast. We need stronger guardrails for data sovereignty and transparency, meaningful standards for model auditing and robustness, and public-sector investment in digital literacy and defensive capabilities. Education and reskilling are not optional: they are the first line of defence against both economic displacement and manipulative uses of digital tools.
Davos 2026 reminded me that influence happens in layers — what’s said on stage matters, but what’s decided in private shapes outcomes. As we race to scale AI and other technologies, we must design institutions that mitigate jaggedness and the concentrated digital power it enables. Otherwise, the peaks of capability will continue to obscure dangerous valleys beneath them.
