The constraints on this data are more interesting than the data, because they decide what a product built on it is even allowed to look like.
A request point rather than a data source
ICANN's Centralized Zone Data Service is where a researcher asks for access. It does not hold the zones. Each request is routed to the registry operator that runs that top-level domain, and that operator decides. ICANN provides the counter, not the goods.
Approval comes one TLD at a time
There is no plan that unlocks everything. Access to .com is one decision by one operator; access to .shop is another decision by another. Coverage of 1,075 generic top-level domains is therefore an accumulation of individual approvals rather than a subscription level — which is also why it grows unevenly and why some namespaces will never be in it at all.
What the agreement forbids
Two restrictions shape everything downstream:
- No substantial portion of a zone may be handed on. Access is for research and analysis, not for redistribution.
- Registries may not be queried at scale. Which rules out resolving a zone-file ambiguity with a WHOIS or RDAP lookup, however convenient that would be.
How a constraint becomes a product shape
A policy note saying “do not request too much” would put the restriction on the operator. Instead every one of the six Actors is built with no input that could express such a request. A brand sweep needs a brand. An availability screen needs a candidate list. A monitor needs a domain list or a filter, and the look-back window is capped at 90 days with no “since forever” option to reach for.
Attribution, and what ICANN does not do
The data is sourced from the Registry Operators' own zone files, obtained through ICANN's Centralized Zone Data Service under agreement with those operators. ICANN does not endorse, sponsor or review any Actor built on it. Access under an agreement is not approval of the product, and a tool implying otherwise is misrepresenting the arrangement.



