LACNIC withdraws 80% ROA resource gate after no consensus

LACNIC’s 80% ROA gate failed consensus. Two days later, the gate was gone.

Why it matters

A proposal would have made route-origin paperwork a condition for getting more addresses. The public record now shows two decisions, two days apart.

Theo March9 November 2026Reading Time 2 minLast update: 9 November 2026

LACNIC’s proposal to make 80% valid ROA coverage an additional-resource condition failed consensus on 4 November and was withdrawn on 6 November.

Summary

LACNIC’s LAC-2026-3 would have required valid ROAs covering 80% of an applicant’s LACNIC-registered IPv4 and IPv6 resources before additional resources could be requested.
The official policy history records no consensus on 4 November 2026 and withdrawal on 6 November. No 80% allocation condition entered force.

The proposed lock was elegant on paper: show valid route-origin authorisations for four-fifths of your existing addresses, then ask LACNIC for more. It put a routing-security measurement at the resource counter. That counter will keep working without the new lock. LACNIC’s policy system says LAC-2026-3 failed its first consensus assessment on 4 November and was withdrawn two days later.

The dates matter. “No consensus” was the assessment; “withdrawn” is the current status. Treating them as interchangeable would erase the choice between another draft and ending this one. LACNIC’s policy development process gives the chairs and author that choice after a failed assessment. The register records the second outcome, but it gives no public account of the objections or the reason for withdrawal.

The proposal, authored by Sergio Kleiman Schwarzstein, sought valid ROA coverage for 80% of the IPv4 and IPv6 resources registered to an organisation and allocated or assigned by LACNIC. Its target was subsequent resource requests. A valid ROA states which autonomous system may originate an address prefix; the proposal tried to turn that operational practice into an eligibility test.

That is a large move for a small percentage sign. The registry would have had to judge a security condition before handing over numbering resources. Now the condition remains a proposal that did not pass. LACNIC’s published history identifies the procedural result; it does not establish a verdict against RPKI or identify the precise technical objection.

The next useful record would be a revised proposal spelling out the measurement, exceptions and consequences. For now, the 80% gate has a complete paper trail and no keyholder.

Original sources: LACNIC proposal history; LACNIC policy index; LACNIC policy development process.