News, events & blog

Nordic Domain Days 2026: The Domain Industry Is Getting Quieter, Smarter, and More Complicated

Where domain conversations used to center on the straightforward relationship between registry, registrar, and registrant, the picture today has a lot more layers. Registry models are shifting, AI is creating new use cases and new demand, certificate lifecycles are getting shorter, DNS abuse needs better measurement, and regulation is putting more pressure on the whole industry than it has in years.

From Thick Registries to Thin Ones

One of the bigger discussions was about the difference between thick and thin registry models. In a thick registry, the registry itself holds the registrant's data. In a thin model, the registry mostly holds a pointer to the registrar, and the actual data sits with them.

Regulation is pushing things toward the thin model. Data minimization is part of it: you're only supposed to hold what you actually need. GDPR and NIS2 have made storing data more expensive and riskier. When both the registrar and the registry are holding the same data, both of them have to deal with quality, security, and compliance. In a thin model, most of that burden shifts onto the registrar.

But there's a catch. When law enforcement or a regulator wants to know which domains are tied to a particular person or organization, the fragmentation that comes with a thin model makes investigating abuse slower. The open question is how to protect privacy while keeping enough visibility that trust in internet infrastructure doesn't erode.

ccTLDs and gTLDs are pulling in different directions here. For profit-driven gTLDs, the thin model makes economic sense. For ccTLDs, which carry more of a public-interest role, data quality is part of what it means to run trustworthy infrastructure. Even if that means taking on more cost and more responsibility.

The Domain Market Is Still Growing

Looking at domain market overviews, one thing stood out: demand hasn't gone anywhere. If anything, AI and new digital workflows are driving more domain adoption, helped along by AI-powered tools that registrars now offer for building websites, generating content, and shaping business ideas.

That's also feeding into higher renewal rates. Once a domain is actually put to use and makes it through that first renewal, the odds of it sticking around go up significantly. A domain's value doesn't come from the moment it's registered - it comes from use, from the service built on top of it, from whether it turns into a real tool instead of staying just an idea. That's part of why discussions about the next round of new gTLDs weren't just about how many might launch, but about what use cases, communities, and business models will actually form around them. And how registrars, registries, and brands are getting ready for that.

AI as a New Layer of Infrastructure

AI's impact is already visible: new naming ideas, content generation, automated workflows, domain-investing tools. But this is still early days. Compared side by side with the traditional web, AI-driven tools can still look like a small slice of the picture. Plot it on a logarithmic scale, though, and the trend becomes obvious: growth is fast, and the effects are compounding.

The AI conversation wasn't only about software, either. Hardware and supply-chain concentration came up a lot. Underneath the AI boom sits a layer end users never see: GPUs, manufacturing capacity, specialized equipment, and a surprisingly small number of companies that most of that infrastructure runs through. That means AI's trajectory is shaped not just by models, but by geopolitics, manufacturing capacity, and physical infrastructure bottlenecks.

There was also a bigger question hanging over the room: the next wave of "internet users" won't just be people, they'll be agents, bots, and automated systems. How do naming, trust, authentication, abuse detection, and accountability work in a world where a large share of activity happens machine-to-machine?

Domain Investing Has Grown Up

Domain investors made the point that success today isn't about spotting a good name. It's about having a systematic workflow. Outbound thinking starts at the moment of registration: who might need this name, which company or industry would benefit from it, and how do you go about demonstrating that interest.

A masterclass on AI tools showed how domain valuation can be broken down into smaller, more manageable tools and skills: RDAP lookups, TLD-fit checks, keyword analysis, trend data, LinkedIn and Wikipedia background checks, trademark screening, SSL history, and comparable sales, all feeding into a single evaluation pipeline. None of this needs to be built from scratch; plenty of ready-made tools already exist. The real skill is in connecting those data sources to an investor's actual strategy: how to value a short .com, which risks to avoid, which past offers worked, and what a portfolio's history has to teach. An "agent memory" of sorts lets that learning carry forward - pulling on an owner's past portfolio, purchase prices, offers, sales, and reasoning - so accumulated experience can scale instead of resetting with every deal.

The Takeaway

The core question coming out of Nordic Domain Days 2026 isn't "are domains still relevant", they clearly are. The real question is who will be able to carry that relevance forward in a changed landscape. That means more regulation, more automation, more AI, and higher expectations around trust. Registries and registrars alike need a clear answer to a simple question: what data are they holding, and why are they holding it?