Summary

  • LAC-2026-5 version 1 would make continued LACNIC services for legacy resources depend on four unlike conditions: formalisation and right-of-use evidence, being “announced and used”, reverse DNS, and current Whois or NIR data plus abuse-c.
  • The proposal remains in discussion and says the Board may activate recovery procedures when the requirements are not met. It does not define how operational use is observed, what exceptions count or how a failure is cured and reviewed.

One line in LACNIC’s latest legacy-resource draft carries more weight than its length suggests. After listing four conditions for continued services, it says that if the requirements are not met, the Board may activate recovery procedures.

The verb matters: may, not must. The status matters too. LAC-2026-5 version 1 was displayed as “En discusión” when captured, with a first consensus assessment shown as the next step. It is a proposal, not an adopted rule, an automatic recovery instruction or a finding against any holder.

The more consequential design choice lies before that sentence. Proposed section 7.6 groups four different kinds of proof. A holder must have taken part in the formalisation process initiated in June 2026 and justified its right of use. The resources must be “announced and used”. Reverse-resolution registration must be maintained. Distribution and assignment information in LACNIC Whois or the relevant national registry, together with the abuse-c contact, must remain current.

These are not four readings of the same instrument. The first is a documentary and procedural record. The third is a DNS state. The fourth is a registry-data state. The second is an operational claim whose two words may not even share an observation surface.

“Announced” can be tested against a chosen set of routing collectors, but any result is bounded by time, vantage point and scope. A prefix can be visible last month and quiet today. It can be announced only to selected peers, carried inside a private network, held for failover or used during an intermittent service. None of those possibilities proves that a particular case satisfies the draft. They show why silence in one public feed cannot, by itself, prove absence of use.

“Used” is wider still. The proposal does not say whether it means routed on the public Internet, assigned inside an organisation, reserved for operational continuity, attached to live infrastructure, or evidenced by some other service. Nor does it say whether “announced and used” is one combined test, two tests that must both pass, or a phrase whose exceptions the Board would define later.

That ambiguity becomes more important because the four predicates converge on the same doorway. A stale abuse-c, a missing reverse delegation, a contested right-of-use file and a route collector’s empty observation are very different cases. If the output record says only “requirements not met”, the evidence has been compressed before the consequential decision begins.

A predicate-level receipt would keep the cases apart. It would name the exact condition said to have failed; the source, scope and window of the observation; the evidence supplied by the holder; any exception or justification; the notice and cure period; the Board decision that follows; the route for review; and whether service or registry changes can be reversed while a dispute is examined.

That receipt would not decide the policy argument in advance. It would make a later decision reproducible. An operator could see whether a registry field was actually stale. A holder could see which route collectors and dates supported an announcement finding. A reviewer could distinguish an evidentiary gap from a decision that the submitted explanation was insufficient.

The proposal’s rationale supplies context but not those mechanics. It says the Board opened a six-month formalisation process on 16 June 2026 to update information and support legitimate, secure use. It lists possible concerns, including resources that cannot be justified, appear unused, have lost organisational continuity or contacts, or may have been usurped. Those are reasons advanced by the proposal’s author, not case findings.

The public proposal record also shows no implementation timetable, references or presentation entry. That does not mean procedures could not be developed. It means the version now open for discussion asks the community to assess a consequence before its observation, exception and review grammar is visible in the text.

The narrow question is therefore not whether LACNIC should care about accurate records or unexplained legacy resources. It is whether four unlike facts should be allowed to collapse into one recovery trigger without leaving a predicate-by-predicate decision trail. A recovery process can be discretionary. Its evidence should not be anonymous.

Source