August 13, 2026
Can One Domain Work in Both the DNS and an Alternative Naming System?
At the registry level, technically yes—but only if “the same domain” means more than the same letters.
It is easy to reproduce a domain-shaped name in two naming systems. The difficult part is preventing that name from becoming two separate identities controlled by different parties.
On August 10, 2026, an ICANN-convened Technical Study Group published an initial draft report offering a preliminary answer. A generic top-level domain could be integrated with an alternative naming system, the group concludes, provided that the same string always remains under the control of the same party and strict operational requirements are met.
This is not approval for any registry to offer the service. The report is a draft, and its first public-comment period remains open until September 21, 2026, at 23:59 UTC. Individual registry services would still have to undergo the applicable ICANN evaluation and receive the necessary contractual authorization.
Two systems, one visible name
The global Domain Name System is the system behind conventional domains used for websites, email, and other Internet services. An alternative naming system could be a blockchain registry, a distributed ledger, or an application-specific namespace.
These systems do not necessarily answer the same kinds of questions. The DNS might associate a name with a website or mail server, while an alternative system might associate the matching name with a wallet address, account, or content identifier.
Nor do they necessarily use the same resolution technology. For blockchain naming systems specifically, ICANN’s technical introduction says resolution is not standardized: implementations may use web APIs, bespoke protocols, or browser plugins.
Consequently, matching spelling does not make two names equivalent. Nor does it make the alternative name resolve through the public DNS or work automatically on a user’s device. Depending on the system, a bridge, dedicated application support, browser plugin, or resolver configuration may still be required.
ICANN’s proposed principle: same string, same controller
The draft calls its model “string+controller integration.”
Suppose the registry operator of a hypothetical .example top-level domain integrates its DNS registry with another naming system. The registry operator would have to control .example in both systems. Alice then registers alice.example.
Under the proposed model, the corresponding name in the alternative system must either be allocated to Alice or withheld exclusively for her. It could not be issued separately to Bob.
Alice would not have to use the name in both systems. She could operate a conventional website through the DNS while leaving the alternative version inactive. Alternatively, she could activate the name in both systems.
The central requirement is coordinated control, not simultaneous use.
“Same string” is also more exact than “looks the same.” Names must match according to DNS rules for labels, capitalization, and internationalized domain names. Visually similar Unicode characters would not be sufficient.
Registration is the easy part
The real test begins after registration.
If control of alice.example passes to a new registrant, control of its alternative-system counterpart must pass with it. A transfer between registrars must likewise preserve the link. The two representations cannot be allowed to acquire different controllers during either process.
Expiration is more complicated. Suppose Alice stops renewing the DNS registration but continues to use the alternative name. The draft says the DNS version would have to remain withheld. The registry could not release it to a new DNS registrant while Alice still controlled the alternative version.
Only when the name has been disabled and deallocated across all the integrated systems could it become generally available again.
Under the draft’s specialized definition, a “suspension” blocks administrative changes without necessarily stopping resolution, and it must apply simultaneously across all integrated systems. If a name were deactivated for an external reason such as abuse or a court order, the registry would have to disable it throughout the integration. The report distinguishes those actions from a routine clientHold, which may be used for a narrower DNS purpose.
An active IETF DNSOP Internet-Draft—a work in progress, not a final standard—similarly identifies lifecycle events, domain-control validation, and synchronization mechanisms as central considerations for responsible DNS integrations.
The danger of “zombie” connections
This lifecycle problem is not merely theoretical.
A May 2026 research preprint examined integrations in which control of a DNS domain was verified when a connection was created but was not reliably checked afterward. The researchers called persistent mappings associated with expired or transferred domains “zombie linkages.”
Among the systems they measured, the authors reported that roughly 24 percent of the examined ENS on-chain imports were zombie linkages. Because the paper remains a preprint under submission, that percentage should be treated as the authors’ measurement rather than an established industry statistic. The underlying design problem is nevertheless clear: proving control once does not prove control forever.
An integration must follow the complete life of a domain—registration, renewal, transfer, expiration, and deletion.
One dependable source of truth
The systems therefore need one dependable, auditable way to determine who controls the name.
The ICANN report describes two broad approaches.
The more straightforward approach would use the conventional registry’s Shared Registration System as the coordinating source. Changes would pass through the established registry infrastructure before being published into the DNS and the alternative system.
A distributed implementation could instead maintain what the report calls a Unified Source of Truth. Updates might involve pending states while a blockchain or another external system settles, but the design would still have to prevent conflicting control.
The report sketches cryptographic tokens as one possible way to demonstrate control. It also says that such a model would still have to meet applicable ICANN contractual requirements and consensus policies, including requirements involving registration data, domain-name disputes, and rapid suspension.
The difficult engineering lies in transfers, outages, and failures—not in creating the first matching record.
What happens when the integration fails?
A credible integration also needs a way to stop safely.
ICANN’s Emergency Back-End Registry Operator program is intended to preserve specified critical registry functions when a registry encounters serious operational problems. The draft observes that an alternative-system integration appears to fall outside those critical functions and therefore may not survive emergency back-end operation.
The draft therefore recommends requiring an applicant to provide a turn-down plan explaining how the integration could be discontinued if it became nonviable. Such a plan would need to address how coordinated control would be preserved while the service was being ended.
Other questions that a proposed service would have to address include protection against network splits, data-integrity assurances for alternative-name lookups—including whether DNSSEC-like protection is needed—internationalized-name processing, anti-abuse procedures, and the application of policies such as the Uniform Domain Name Dispute Resolution Policy and Uniform Rapid Suspension System.
What this would not mean
The integration would not make the alternative system part of the public DNS or delegate its separate namespace in the DNS root. It would not make every blockchain or alternative name universally accessible, and it would not allow an individual registrant to create a genuine registry-level integration simply by registering or minting the same spelling somewhere else.
The model is a registry-level service. An existing registry operator would have to seek approval individually through ICANN’s Registry Services Evaluation Policy process; a prospective registry could be evaluated as part of a New gTLD Program application. A technical-panel review may be required, and an approved service would need corresponding registry-agreement language.
What would this mean for SEO?
Nothing in the ICANN report or Google Search documentation establishes a search-ranking advantage from the integration itself. The integration alone does not create another indexable web page. Google’s minimum technical requirements apply to pages, not naming-system registrations: Googlebot must be able to access a page, receive an HTTP 200 response, and find indexable content.
If equivalent content is available at more than one conventional web URL, Google may group the pages as duplicates and select a canonical. Publishers can indicate a preference with redirects, rel="canonical" annotations, and sitemap inclusion, and should link internally to the preferred URL. These are signals rather than guarantees, and Google describes redirects and canonical annotations as stronger signals than sitemap inclusion.
One identity rather than two similar labels
Can one domain work in both the DNS and an alternative naming system? Technically, yes.
But the achievement is not duplicating the label. It is maintaining one unambiguous controller through every registration, transfer, expiration, suspension, failure, and dispute.
This makes the proposal less a story about replacing the DNS than about extending a familiar name into another environment without allowing its identity to split.
Without those guarantees, the result is not one domain working in two systems. It is two different names that happen to look alike.
Principal sources