GDC 2026 showed how identity, payments, agents and assets are increasingly converging around the same infrastructure questions, with naming emerging as a key layer for continuity and interoperability onchain.
Earlier this month, I spent three days at the Global Digital Collaboration Conference (GDC) in Geneva. It was only the second edition, but it brought together an unusual mix of people working in governments, standards organisations, open-source projects and some of the world's largest technology and payments companies.
The program covered digital passports and cross-border identity, AI agents making payments, organisational credentials, privacy, stablecoins and digital wallet standards, with major technology companies, standards bodies and public institutions all represented.
Even though the topics seemed separate at first, after three days I kept coming back to many of the same questions.
How do you identify a person, organization, asset or agent within different systems? Who gets to decide what is trusted? What happens when wallets, keys or credentials change? And how do any of these systems work without depending on one company or one closed platform?
Most of these questions existed long before Ethereum and ENS. The questions remain the same, but how they need to be answered has changed. Payments are moving onchain. Assets are being tokenized. AI agents are beginning to transact.
At GDC, it became increasingly difficult to see these developments separately.
One question raised during GDC was where an identity actually lives. If a credential is stored in a wallet, does that make the identity dependent on the wallet or its blockchain address? The problem becomes obvious when an address changes or the credential expires.
The discussion didn't lead to one clear answer, but it highlighted that an identity, its credentials and the infrastructure used to manage them are not necessarily the same thing. A company remains the same legal entity even as its credentials change. Similarly, a person doesn't become a different person when their passport expires.
As more identity infrastructure becomes digital and potentially onchain, continuity becomes increasingly important. If credentials, wallets, addresses and keys can all change, what should remain persistent?
These questions become more complicated when an AI agent acts on someone's behalf. Agentic commerce was a recurring topic at GDC. Today's payment and authentication systems were designed around a person approving a transaction. Agents may need their own identifiers and permissions while remaining connected to the person or organisation they represent.
This is where conversations that initially seemed unrelated began to overlap. A traveler presenting a digital credential, a company operating on different financial networks, an AI agent purchasing on someone's behalf and a tokenized asset linked to an issuer have little in common at the application level. Yet each depends on identity, authority and a reliable way to refer to the right entity.
That last requirement is easy to underestimate because naming tends to disappear once it works. We rarely think about the domain when opening a website. But as more activity moves onchain, how we refer to it consistently becomes less of a user-experience problem and more of an infrastructure problem.
ENS began by solving the most visible version of this problem. Raw blockchain addresses are useful for machines and unsuited for humans. A name like enslabs.eth gives users a persistent, readable way to interact with an address that could otherwise change underneath it. It may have been the simplest expression of a much broader use case.
The internet offers a fitting precedent. DNS did not only become foundational because domains were easier to type than IP addresses. Its deeper value was the abstraction it provided. Servers could move, IP addresses could change and new services could emerge without requiring people to learn a new way of finding them.
Something similar can now be observed onchain. A person can have several wallets, an organization can operate on multiple networks, and an AI agent may need an identifier and payment address. These are different use cases, yet all benefit from a stable reference point while the infrastructure around them changes.
The range of use cases can make ENS appear to be moving into several different categories at once. But ENS does not need to become a payment network, identity provider or RWA platform to support any of them. An entity or community can operate its own namespace powered by ENS and define what the names beneath it represent, without requiring ENS itself to determine what those entities are or whether they should be trusted.
Good infrastructure doesn't need to predict every application that will depend on it. DNS didn't need to anticipate streaming or cloud computing to become useful to both. ENS may be following a similar path. Its versatility may say less about becoming many different things than about what happens to naming once more infrastructure moves onchain.
One of the most interesting things about GDC was how many technical communities were trying to solve adjacent versions of the same problem, from platform companies and the Ethereum Foundation to standards bodies, governments and open-source communities. The future being discussed in Geneva didn't look like one identity system winning.
In some areas, the tech is already surprisingly mature. ZK can prove attributes without revealing the underlying data, while the EF's work on unlinkability aims to prevent repeated use of a credential from automatically connecting separate interactions. Other work focuses on key rotation, allowing an identity to survive when the cryptographic keys behind it change. Existing infrastructure is also being adapted, including attempts to bridge newer decentralised identifiers with X.509 systems already used by governments and regulated industries.
At the same time, there are hundreds of DID methods, several credential formats and competing approaches to wallets, trust registries and agent identity. That may look like fragmentation, but in a field this early it can also mean we simply have not decided which solutions deserve to become permanent.
The ecosystem isn't waiting for one universal solution. The future looks more like a network of different systems that will need to recognize and interact with one another.
That is why interoperability becomes more than a standards or governance problem. It also becomes a naming problem.
For ENS, this may be where several of these threads begin to meet. The same ENS name used to connect to an asset address could also provide a persistent reference for credentials or an agent. A credential wallet and an asset wallet don't need to become the same system for both to connect to the same name. ENS can provide continuity across networks and applications while the infrastructure behind that name changes.
That continuity does not require every credential or attribute to become public. ZK proofs, selective disclosure and other privacy-preserving systems can prove what is necessary while keeping the underlying information elsewhere.
GDC was only in its second year, and much of the technology discussed there is still being defined. That uncertainty is valuable because the architecture has not hardened yet and the boundaries between digital and onchain infrastructure are still being negotiated.
That may ultimately be the most important takeaway from GDC: the questions are far from settled, but the communities trying to answer them are no longer working in isolation.