ARIN 57 Policy Proposals — Summary & Client Impact

April 20, 2026

ARIN 57 runs from 19-22 April 2026 in Louisville, with the policy block scheduled for Tuesday, 21 April. This round features six draft policies focused on cleaning up edge-case ambiguity in the NRPM, making IPv6 policy easier to apply in practice, and tightening the relationship between policy text and how staff already operate. The final policy takes a turn with a focus on extraterrestrial networks.

None of the first five proposals looks like a market shock; for operators, brokers, and address strategy teams, the most useful policy changes are often the ones that remove interpretation risk, shorten request cycles, and make it easier to predict how ARIN will read a request before submission. ARIN’s own framing of the docket leans heavily in that direction: these are mostly cleanup, clarification, and consistency proposals, with one genuinely forward-looking exception in ARIN-2026-1.

ARIN-2025-1: sorting out ISP versus LIR

One of the long-standing issues in ARIN’s Number Resource Policy Manual (NRPM) is the inconsistent use of the terms Internet Service Provider (ISP) and Local Internet Registry (LIR). This is the sort of proposal may not be exciting but in practice has real value. ARIN says the current text uses “ISP” and “LIR” somewhat interchangeably, even though only LIR is formally defined in Section 2 today. The proposal would define both terms and update the NRPM to use “LIR” where appropriate, bringing ARIN’s language closer to the terminology already used by the other RIRs.

For clients, especially organizations operating across more than one RIR region, this is a good change even if it does not alter allocation outcomes. It lowers the odds of category confusion during requests, trims some of the translation work global teams have been doing internally, and makes ARIN policy read more like the rest of the RIR world. In other words, this is not a policy that changes who gets space. It is a policy that makes the rules easier to parse before a request turns into a debate over vocabulary.

ARIN-2025-3: a smaller in-region footprint for out-of-region IPv4 use

ARIN’s current rule allows out-of-region use of ARIN-issued IPv4 space only if the organization is already using at least a /22, or equivalent aggregate, inside the ARIN region. Draft Policy ARIN-2025-3 would lower that minimum to a /24. ARIN’s own explanation is blunt: the /22 threshold can harm smaller organizations that need some global deployment flexibility without having a large North American footprint.

Thisitem is very meaningful for smaller networks. A /24 is a far more reachable baseline for startups, edge-heavy platforms, and distributed operators that may have customers or infrastructure tied to the ARIN region without concentrating enough in-region usage to satisfy a /22 test. The practical effect is more flexibility with less artificial architecture. It will not erase the regional tie altogether, but it does make compliance with Section 9 feel more proportionate to how modern networks are actually built.

ARIN-2025-6: fixing the IPv6 math

Addressing a section where a quiet formula error can create an outsized amount of friction, ARIN-2025-6 corrects the equation used in Section 6.5.2, which covers initial IPv6 allocations to LIRs. It also updates the older term “Provider Assignment Unit” to “Provider Allocation Unit” so the text matches current practice.

This change holds a lot of operational value. When the formula in the text and the intended outcome are not perfectly aligned, applicants can over justify, under justify, or burn cycles trying to reverse-engineer what staff will accept. This draft policy will not reshape IPv6 planning, but it should reduce avoidable confusion and make sizing discussions cleaner for both seasoned applicants and teams that are still early in their IPv6 deployment maturity.

ARIN-2025-7: making the single-site /48 explicit

ARIN says this proposal, titled “Make Policy in 6.5.8.2 Match the Examples,” is intended to remove any doubt that a single-site organization qualifies for an IPv6 /48 allocation. That is a modest edit, but it closes a gap between how people read the examples and how they read the actual operative text.

That matters because small end-user organizations are often where IPv6 policy ambiguity becomes unnecessarily expensive. If a single-site enterprise, institution, or smaller operator has to spend extra time proving what the examples already imply, the policy can be clearer. Codifying the /48 outcome should cut down on back-and-forth, speed up approvals, and make IPv6 feel less like a bespoke exercise for smaller deployments.

ARIN-2025-8: closing the loop on Section 4.10

Section 4.10 is the special IPv4 space reserved to facilitate IPv6 deployment. ARIN says the current text does not explicitly bar out-of-region use beyond the general Section 9 framework, even though allowing that use appears inconsistent with the original intent of the reserved pool. ARIN-2025-8 would make the in-region restriction explicit, and ARIN notes that this aligns with current staff implementation already.

For most clients this is more of a policy-text catch-up exercise, but it does matter because it removes the idea that Section 4.10 might be stretched into a globally deployable workaround. If you are using reserved IPv4 transition space, ARIN wants the rulebook to say plainly what staff have already been enforcing in practice. From a stewardship perspective, that is hard to argue with. From a planning perspective, it is better to know the lane is closed than to design around a loophole that was never really open.

ARIN-2026-1 TIPTOP

TIPTOP, short for Taking IP To Other Planets, comes out of a real problem ARIN has now put into writing: organizations building IP-based networks beyond Earth’s orbit are already using address space from multiple RIRs without coordination, and that can create routing scalability concerns. The policy is being developed jointly with the IETF TIPTOP working group and would create a framework for IPv4 and IPv6 allocations for networking infrastructure in outer space, beyond geostationary orbit, organized hierarchically around celestial regions. It also introduces the term Extra-Terrestrial Network, or ETN.

For satellite operators, aerospace programs, and anyone thinking seriously about lunar or deep-space networking, the value is immediate: legitimacy, structure, and less risk of ad hoc addressing decisions becoming long-term routing headaches. For everyone else, TIPTOP is still worth watching because it shows ARIN trying to get ahead of a technical problem before the market forces a messy workaround. It is also a reminder that Internet numbering policy, for all its reputation as plumbing, still has room for a little science fiction becoming operations reality.

What clients should really take away from ARIN 57

Beyond the novelty of TIPTOP, the core theme of ARIN 57 is policy legibility. The docket is not about rewriting the foundations of number resource management. It is about making the NRPM easier to interpret, more aligned with existing practice, and better suited to the realities of smaller, globally distributed, and IPv6-building organizations. That may not make headlines the way a major transfer-policy fight would, but for clients navigating requests, utilization, and long-term address planning, these are the changes that reduce friction where it actually hurts.

The best way to think about this meeting is simple: one proposal makes ARIN’s terminology more globally coherent, one makes out-of-region IPv4 use less punishing for smaller operators, two make IPv6 policy easier to apply correctly, one seals up a transition-space edge case, and one asks what IP addressing should look like when the network is no longer on Earth. That is a pretty good ARIN agenda: practical, procedural, and just ambitious enough to remind everyone that numbering policy still has a future-facing edge.