The review meeting begins with a spreadsheet that appears to be about progress. An African operator has more IPv6-capable routers than it had three years ago. Its mobile core can carry IPv6. Its enterprise firewalls can be upgraded. A cloud migration partner says new workloads should be IPv6-ready by default. The regulator wants evidence of a national transition plan. Procurement has collected vendor quotations. The engineering team has a slide showing the share of traffic that could move over IPv6 if enough customers, banks, public agencies, foreign partners, payment processors, monitoring platforms and legacy devices followed.
Then the commercial director asks a less comfortable question. Which contracts can be delivered if the company assumes IPv4 demand disappears in the next budget cycle? The room slows down. The bank customer still wants stable IPv4 endpoints for fraud systems and partner allowlists. The public-sector buyer still has older systems and regional counterparties. The managed-security client still wants predictable egress. The data-centre tenant wants public addressing that upstreams, geolocation vendors, mail providers and auditors understand. The mobile business can expand IPv6 on handsets, but it still needs IPv4 reachability for applications, customer support, attribution and interconnection. Nobody in the room opposes IPv6. Nobody can run the business as if IPv4 has become irrelevant.
That is the political economy of IPv6 transition. The transition is real, but it is not a replacement event. It is a long period in which a new protocol grows while the old one remains economically decisive. During that period, the scarcity of IPv4 does not vanish. Registry records do not become less important. Transfer, leasing, routing evidence, reverse DNS, resource certification and customer due diligence remain part of the operating surface. The future protocol can reduce some future address pressure, but it cannot retroactively erase installed systems, counterparty habits, procurement clauses, old devices, national institutions, cloud compatibility assumptions or the market value of the scarce identifiers already embedded in business relationships.
AFRINIC matters because it shows what can happen when transition rhetoric is asked to carry more institutional weight than it can bear. AFRINIC is the Regional Internet Registry serving Africa and the Indian Ocean region. It distributes and registers IPv4 addresses, IPv6 prefixes and autonomous system numbers, and supports services that make registry records useful to networks and counterparties. Its own published materials record that the region entered IPv4 Soft Landing Phase 2 on 13 January 2020, with a minimum IPv4 allocation or assignment size of /24 and a maximum of /22 per request under the Phase 2 rules. That is rationing. It is not abundance.
At the same time, AFRINIC's public history contains the kind of institutional stress that makes registry power visible. Public reporting has described a major alleged address misappropriation history, the Cloud Innovation dispute, bank-account freezes in Mauritius, multiple court cases, receivership, years without normal board continuity, election disputes and continuing questions over who can legitimately speak and act for the registry. These facts do not prove that IPv6 is bad or that AFRINIC has no legitimate function. They prove something narrower and more important: the medium-term transition from IPv4 dependence to broader IPv6 deployment cannot be governed by slogans that ask operators to forget the economic and legal reality of IPv4 records.
The central issue is not whether networks should deploy IPv6. They should, where it reduces future constraint and improves technical reach. The issue is whether the story of IPv6 is used to postpone the harder work of IPv4 stewardship: clear records, bounded registry discretion, reliable continuity, credible dispute handling, market-compatible transfer rules and accountability for the ledger that operators, customers and financiers still rely on. A transition programme that treats IPv6 as a moral escape from IPv4 scarcity becomes a political instrument. It lets incumbents and registries invoke tomorrow's protocol to justify today's gatekeeping.
The transition starts as coexistence, not replacement
The first economic distinction is between migration and replacement. A replacement event has a date. A company retires a switch, shuts down a billing system, exits a data centre or completes a customer migration. After that point the old input can be removed from planning. IPv6 has not produced that kind of event for the public Internet. It has produced a long coexistence period in which many networks, devices and services can use IPv6 while a vast quantity of commercial, public-sector and security activity still assumes IPv4 compatibility.
This matters because capital plans like dates. A regulator can ask for an IPv6 transition roadmap. A board can approve a multi-year equipment refresh. A vendor can sell IPv6 readiness. A national strategy can set adoption targets. But the economics of a network are governed by counterparties. If a bank, hospital vendor, customs system, payment processor, enterprise customer, foreign cloud service, monitoring platform or upstream provider still treats IPv4 as the compatibility baseline, the operator cannot retire IPv4 merely because its own equipment supports IPv6.
IPv6 deployment therefore changes the shape of IPv4 demand rather than eliminating it. Some traffic shifts. Some new services are designed more cleanly. Some internal architecture improves. Some future address growth is avoided. But the remaining IPv4 demand becomes higher value precisely because it is tied to compatibility, trust and legacy reach. The addresses that remain necessary are not necessarily numerous in every use case, but they are commercially sensitive. They are used for gateways, customer-facing services, business endpoints, mail reputation, public services, managed security, translation pools, monitoring and customer-specific exceptions.
The economics of this coexistence are often lost in public transition narratives. A chart showing IPv6 capability can make progress look linear. The operating reality is discontinuous. A mobile handset may support IPv6 while an enterprise fraud system still expects IPv4 evidence. A cloud workload may speak IPv6 while a public agency's procurement file names IPv4 whitelisting. A university network may introduce IPv6 while research partners, building systems and legacy applications continue to use IPv4 assumptions. The old protocol retires not when engineers approve it, but when the weakest important counterparty no longer needs it.
That is why IPv4 scarcity remains a medium-term economic fact. The issue is not the theoretical size of IPv6. The issue is the installed economic system around IPv4. Scarcity attaches to a public identifier that market participants already recognize, route, certify, record, diligence, lease, finance and price. A larger replacement namespace can reduce the need for future scarcity, but it does not dissolve existing reliance on the scarce namespace. In other infrastructure sectors, this would be obvious. A new rail gauge, payment standard or energy source does not instantly remove the value of the old network while customers, equipment and contracts still depend on it.
For AFRINIC-region networks, the distinction is sharper because transition coincides with scarcity. The region is not moving from abundant IPv4 to abundant IPv6 in a clean sequence. It is expanding digital services during a period when IPv4 is already rationed and the regional registry has been under legal and governance stress. That combination makes IPv6 a useful technical programme but a dangerous political alibi. If the transition story tells operators that IPv4 no longer deserves serious institutional protection, it asks them to ignore the very input they still need to satisfy customers.
Installed systems give IPv4 a long half-life
The reason IPv4 persists is not nostalgia. It is the installed base. Every network layer has a physical and contractual memory. Routers, firewalls, customer-premises equipment, lawful-intercept systems, billing tools, anti-fraud platforms, logging systems, peering filters, mail systems, DNS habits, operations manuals, public tenders and audit files have been built around IPv4 for decades. Replacing that memory is slower than enabling a new protocol on capable equipment.
Installed bases are powerful because they create coordination costs. One firm can move quickly only if others move with it. A bank may upgrade its own network but still require IPv4 from merchants because thousands of counterparties use old systems. A government may announce IPv6 support but keep IPv4 endpoints for citizens, schools and vendors. A cloud provider may offer IPv6 but still charge, structure or document IPv4 because customers demand it. A small ISP may deploy IPv6 to subscribers while maintaining IPv4 pools for services that break, degrade or require manual exception when translation is too aggressive.
The half-life is also institutional. Procurement departments are conservative because outages are expensive. Security teams prefer known evidence. Auditors prefer stable records. Lawyers prefer continuity. Public agencies prefer compatibility. A tender that requires public IPv4 may be technically out of date, but it can still decide who wins the contract. A bank whitelist may be inelegant, but it can still determine whether a payment integration launches on time. An anti-abuse system may be crude, but it can still block or rate-limit a service. These are not engineering ideals; they are economic constraints.
IPv6 advocates sometimes respond that these habits should change. They may be right. But "should" is not a retirement mechanism. Someone must pay for rewriting the procurement clause, upgrading the firewall, changing the audit checklist, testing partner systems, retraining staff, redesigning logging, adjusting geolocation and proving to customers that the new arrangement is safe. Until that happens, IPv4 remains the shared compatibility language among parties that do not fully control one another.
This is where the transition narrative can mislead. It treats IPv6 as if it were a purely technical remedy to a technical shortage. In practice, the shortage is embedded in a social and economic coordination system. IPv4's value comes not only from packet delivery but from the fact that other parties know how to evaluate it. Registry records, routing evidence, reverse DNS, RPKI status, abuse contacts and historical use all help counterparties decide whether a network is legitimate and stable. Those functions cannot be replaced by the existence of a larger address family alone.
AFRINIC's region includes markets where equipment refresh cycles, public-sector procurement, bank integration, small-business technology and consumer-device turnover do not all move at the same pace. Some parts of the network can be modern. Others remain conservative for good reason: budgets, risk, imported equipment costs, public-service continuity and low tolerance for service failure. The installed base therefore gives IPv4 a long economic half-life, especially for services that touch money, identity, health, education, customs, security and international counterparties.
The result is a transition period in which IPv6 adoption and IPv4 stewardship are complements, not substitutes. A serious policy would say: deploy IPv6 where it reduces future pressure, and at the same time make the IPv4 ledger more reliable because the old layer will remain economically critical. The weaker policy says: IPv6 is the future, so disputes over IPv4 records are legacy noise. AFRINIC's history shows why the weaker policy is unsafe.
Counterparties retire scarcity, not protocol advocates
Scarcity is not retired by declaration. It is retired when the marginal buyer no longer needs the scarce input. In the medium term, the marginal buyer of IPv4-compatible service is not an ideology. It is a bank, public agency, enterprise customer, hosting tenant, mobile subscriber group, security vendor, upstream provider or cross-border partner that has its own systems and risk tolerances.
That is why IPv4 continues to price trust. A customer asking for stable public IPv4 is often asking for more than address space. It is asking whether the provider can be reached by old networks, whether fraud teams can identify traffic, whether emails will arrive, whether APIs can be whitelisted, whether geolocation will be tolerable, whether support staff can troubleshoot, whether the provider can survive a registry update and whether the address will remain attached to the service long enough to justify integration. The public address becomes a promise of operational continuity.
IPv6 can support many of these functions when both sides are ready. The problem is that readiness is uneven. A network can control its own deployment; it cannot control every partner's deployment. A public-service vendor may be technically advanced but still serve citizens using old devices and networks. A data-centre provider may offer IPv6 racks but still need IPv4 for tenants whose customers have not changed. A payment processor may test IPv6 but keep IPv4 risk controls because the losses from a failed migration exceed the benefits of appearing modern.
The political economy follows from this asymmetry. The party most enthusiastic about IPv6 is not always the party bearing the risk of a failed counterparty transition. A standards advocate may count adoption. A vendor may sell upgrades. A registry may justify continued relevance through IPv6 allocation and training. A government may announce a plan. The operator faces the customer. If the customer's service fails, the operator pays through credits, churn, reputation, support burden and lost contracts.
This does not make operators anti-IPv6. It makes them economically rational. They will deploy IPv6 when the expected benefits exceed the costs and risks. They will keep IPv4 where counterparties still demand it. The transition is therefore endogenous to business relationships. It cannot be reduced to a moral distinction between progressive IPv6 adopters and backward IPv4 holders.
AFRINIC's Phase 2 scarcity makes the point concrete. If a network can receive at most a small IPv4 allocation or assignment under the remaining pool rules, it must decide which customers and services receive scarce public identity. IPv6 may reduce some growth pressure, but the scarce IPv4 units that remain available become strategic. They are allocated internally to the services for which counterparties still require compatibility. The registry's role in recognizing and maintaining those units therefore remains powerful.
A responsible transition policy would treat counterparties as the binding constraint. It would ask which sectors still require IPv4, why they require it, what evidence would let them accept IPv6, how long the conversion will take and how to protect IPv4-dependent services while that conversion proceeds. A political transition narrative does something easier. It assumes the future protocol carries enough moral force to discipline present behavior. That assumption leaves operators exposed.
Compatibility bridges show where trust still sits
Network Address Translation, carrier-grade translation and other compatibility mechanisms are often presented as evidence that the IPv4 problem can be engineered around. They can help. They also prove the opposite of the strongest transition claims. Translation exists because networks still need to reach an IPv4 Internet. It is a bridge built for dependence, not evidence that dependence has disappeared.
The economics of translation should be kept distinct from a full account of who pays each operational cost. That is a separate incidence question. Here the point is narrower. Translation is a market signal. If operators, cloud providers, enterprises and mobile networks invest in compatibility layers, they are revealing that IPv4 reachability remains valuable. If a network must maintain shared egress, public pools, logs, exception lists and customer-specific endpoints, it is revealing that the old addressing layer still carries commercial trust.
Translation also changes the character of demand. A household subscriber may not need a dedicated public IPv4 address. A banking integration may. A public API may. A mail stream may. A managed firewall may. A hosting tenant may. A payment endpoint may. A monitoring system may. A customer with compliance needs may. The total number of public addresses can be reduced, but the addresses that remain are attached to higher-stakes functions. That makes their registry status more important, not less.
In an abundant world, an address error is often administrative. In a scarce, translated, reputation-sensitive world, an address error can be commercial. If a public egress address is listed incorrectly, if reverse DNS is not maintained, if routing evidence is stale, if a registry dispute delays a contact update, if a transfer is uncertain, or if a block carries unresolved history, the address can lose value even while packets move. Translation concentrates more customers and services behind fewer public identifiers; that concentration increases the importance of clean, stable, trusted public identifiers.
This is why the language of "just deploy IPv6" can become a way of avoiding ledger questions. It suggests that the answer to scarcity is a future architecture, while the current business still depends on the credibility of IPv4 records. A provider that tells a bank "we are moving to IPv6" has not answered whether the bank's existing IPv4 allowlists, fraud controls, logs and partner systems will keep working next quarter. A registry that says IPv6 is the long-term answer has not answered whether its IPv4 records are accurate, portable and protected from discretionary overreach.
AFRINIC's alleged address-theft history makes this point unavoidable. KrebsOnSecurity reported in 2019 on allegations that a senior AFRINIC figure had been linked to companies involved in selling African IPv4 blocks, with researcher Ron Guilmette estimating the market value of documented affected addresses at more than US$50 million. Internet Governance Project later connected that history to the wider crisis, describing how prior misappropriation allegations set the context for aggressive enforcement and the Cloud Innovation dispute. Whatever the final legal treatment of particular allegations, the economic lesson is plain: when IPv4 has market value, the integrity of registry records becomes market infrastructure.
IPv6 deployment does not erase that integrity problem. It may reduce some future pressure to steal, lease, hoard or fight over IPv4. But during the transition, the remaining IPv4 records become more valuable precisely because they bridge old and new networks. Translation is the visible proof. A network that still needs bridges still needs reliable records for the land on both sides of the bridge.
Phase 2 scarcity turns adoption into politics
AFRINIC's own exhaustion materials supply a narrow factual anchor. Phase 1 began in 2017. Phase 2 began on 13 January 2020. Under Phase 2, the minimum IPv4 allocation or assignment is /24 and the maximum is /22 per allocation or assignment. Applications are submitted through tickets, complete applications proceed to evaluation, and additional requests require evidence of efficient use of previously delegated AFRINIC IPv4 space. These details are administrative. Their effect is political economy.
Rationing changes power. When the free pool is abundant, a registry's gatekeeping has a different character. Applicants may dislike delay, but the resource is not yet a high-value asset in the same way. When the pool is narrow and each allocation is small, a registry's decisions, documentation standards, timing, good-standing requirements and interpretation of need affect who can grow, who must lease, who must buy, who must depend on upstream addresses and who must redesign services around scarcity.
The transition story can hide this shift. If IPv6 is framed as the solution that makes IPv4 scarcity obsolete, then Phase 2 looks like a temporary inconvenience on the way to abundance. But for operators making medium-term plans, Phase 2 is not temporary enough to ignore. It defines today's bargaining position. It affects today's customer commitments. It shapes today's balance between leasing, transfers, translation, conservation and IPv6 deployment. A future reduction in IPv4 dependence does not help a current contract that requires IPv4-compatible delivery.
Scarcity also changes the registry's legitimacy problem. In the allocation era, a registry could defend itself as a steward distributing a low-priced public resource according to need. In the post-exhaustion era, much of the economically important action shifts to record maintenance, transfers, disputes, routing evidence and continuity of previously allocated resources. The registry's legitimacy should therefore become more ledger-like: accurate records, predictable updates, narrow discretion, transparent correction, secure publication and reliable preservation of running networks. If it instead keeps an allocation-era posture of broad stewardship, it risks turning scarcity into institutional leverage.
This is where IPv6 rhetoric can serve gatekeeping. The registry can say it is promoting the future while retaining control over the scarce present. It can portray IPv4 market behavior as backward or suspect while its own records remain indispensable to any market transaction. It can invoke conservation and transition while avoiding a candid account of rights, remedies, liability and portability. The old pool may be exhausted, but the institutional habit of permission can continue.
AFRINIC's crisis is a test case because its Phase 2 scarcity coincided with contested institutional authority. A small remaining pool, a history of record-integrity concerns, a dispute over resource use, court-ordered freezes, receivership and board-election uncertainty all sat on top of the same medium-term reality: African networks still need IPv4 compatibility while trying to deploy IPv6. In that environment, transition language is not neutral. It can either clarify the coexistence problem or distract from it.
The better reading is sober. IPv6 is a necessary long-run investment. IPv4 remains a scarce medium-term production input. The registry function remains valuable because scarcity needs clean records. Any policy that collapses those three statements into a slogan will misprice risk.
Future protocol promises can become present alibis
Every institution has an incentive to describe its power in terms that sound public-spirited. In the numbering system, IPv6 provides a particularly useful vocabulary. It can be described as inclusion, future-proofing, abundance, technical progress and global responsibility. Much of that language has truth in it. The problem begins when the future protocol becomes an alibi for present discretion over IPv4.
An alibi works by changing the moral frame. If the future is IPv6, then the operator defending IPv4 transferability can be painted as clinging to the past. If abundance is the ideal, then IPv4 pricing can be described as socially suspect even when prices are the mechanism that moves scarce resources toward demand. If stewardship is invoked for transition, then registry control can be portrayed as necessary guardianship rather than as a chokepoint over valuable records. The discussion shifts from institutional accountability to moral posture.
This is analytically weak. A network can believe IPv6 is the future and still require strong IPv4 property-like assurances in the present. A company can deploy IPv6 and still need predictable transfer rules. A regulator can set IPv6 targets and still require the registry to maintain due process around scarce IPv4. A customer can accept future migration and still demand that today's service be reachable by IPv4. These positions are not contradictory. They describe the actual transition.
The danger is that transition rhetoric weakens the discipline that scarcity should impose on registries. Scarce assets demand better records, not vaguer ones. They demand clearer remedies, not symbolic liability. They demand independent dispute handling, not institutional self-help. They demand careful separation between registry maintenance and enforcement. They demand continuity plans that protect networks and customers rather than institutional pride. If the answer to each demand is "deploy IPv6," the registry has used the future to avoid accountability for the present.
Lu Heng's public notes express this point in sharper ideological language, arguing that IPv6 abundance preserves administrative discretion while IPv4 scarcity forces price discovery, accountability and operator leverage. A reader need not accept every conclusion to see the incentive problem. Institutions that administer abundance can keep allocating, training, convening and validating need. Institutions that administer scarcity face pressure to clarify rights, transfers and liability. IPv6 can therefore be both a technical protocol and an institutional comfort.
AFRINIC's public facts make this more than theory. The registry has had to manage not only IPv6 promotion but also a scarce IPv4 pool, address-theft allegations, contested enforcement, litigation, receivership and election legitimacy. In that setting, the transition narrative can be used in two ways. It can honestly tell operators to prepare for long coexistence. Or it can imply that arguments about IPv4 rights are an unhelpful distraction from the future. The second use is politics disguised as engineering.
The credible position is not anti-IPv6. It is anti-evasion. IPv6 adoption should be measured and encouraged on its own merits. IPv4 stewardship should be judged by the quality of the ledger, the fairness of process, the reliability of continuity and the ability of operators to make rational market decisions during the coexistence period. Combining the two lets weak performance in one domain be excused by ambition in the other.
AFRINIC's crisis exposes weak ledger discipline
A registry record is not merely a line in a database once the resource is scarce, routable, transferable, financeable and embedded in contracts. It is evidence used by customers, upstream networks, cloud platforms, auditors, banks, courts, buyers, sellers and operators. Weak ledger discipline therefore has a direct economic cost. It raises diligence expense, increases uncertainty, slows transactions and creates discounts around otherwise usable resources.
AFRINIC's crisis shows the point in several layers. First, the alleged address misappropriation history undermined confidence in record integrity. Second, the Cloud Innovation dispute raised questions about the boundary between legitimate registry review and expansive control over how resources are used. Third, the Mauritius court process and bank freezes showed that disputes over address resources can affect the registry institution itself. Fourth, receivership showed that the registry function had to be preserved even when ordinary governance had failed. Fifth, board-election disputes showed that legitimacy over who acts for the institution is not ceremonial; it affects resource-holder confidence.
None of these layers is solved by IPv6 deployment. A network running IPv6 still needs the registry to maintain accurate IPv4 records for services that remain IPv4-dependent. A bank accepting an IPv6 migration plan still wants assurance that the provider's current IPv4 endpoints will not become entangled in a registry dispute. A buyer of addresses still needs confidence that the seller's records are clean. A lessor still needs the public record to match the contract. A customer still needs reverse DNS and routing evidence to be maintained. A court still needs a ledger that can be understood.
The ledger problem is also distributional. Large operators can afford lawyers, brokers, consultants and technical staff to reconstruct history. Smaller networks cannot. If registry records are unreliable or institutional actions unpredictable, the market advantages those who can buy assurance privately. The public registry then fails in one of its central economic functions: lowering transaction costs for everyone. It becomes one more layer of asymmetric information.
Internet Governance Project's 2021 analysis described AFRINIC as having only ever held a small share of global IPv4 space and framed the Cloud Innovation conflict as a fight over a small remaining resource in a much larger global scarcity problem. That framing is useful because it shows the scale mismatch. The resource pool at issue may be small relative to African future demand, but the institutional conflict around it can still damage confidence in the whole recognition layer. A small pool can create a large risk premium when the ledger is contested.
Ledger discipline means more than uptime. It means the record is accurate, changes are authorized, disputes are marked without unnecessary disruption, routine maintenance continues, fraud is corrected with evidence, transfers are handled predictably and governance conflicts do not spill into running networks. It also means the registry does not use its position as ledger operator to claim broad power over business models unless clear policy and lawful process support the action. The more valuable IPv4 becomes during the transition, the more important this discipline becomes.
If IPv6 rhetoric diverts attention from ledger discipline, it increases rather than reduces transition risk. The path to a less IPv4-dependent Internet passes through years of dependence on high-quality IPv4 records. A weak ledger cannot be excused by a better future namespace.
Receivership separates continuity from prestige
AFRINIC's receivership is often discussed as a dramatic governance event. Economically, its most important lesson is simpler: registry continuity is a function that can and must be separated from institutional prestige. Internet Governance Project reported in October 2023 that AFRINIC had been moved into receivership by order of the Bankruptcy Division of the Supreme Court of Mauritius. The receiver's role was described as maintaining the status quo, preserving the value of the business, overseeing an election process, helping form a proper board and appointing a chief executive under court oversight.
The fact of receivership did not mean African networks no longer needed registry services. It meant those services were important enough to require continuity while ordinary governance was repaired. That distinction should shape the transition debate. The registry function matters: uniqueness, registration accuracy, RDAP and Whois publication, reverse DNS, RPKI, routing records, contact maintenance, dispute notation and resource-holder support. But the importance of those services does not prove that every authority claim made by the incumbent institution is necessary.
This is where transition politics often becomes confused. Defenders of a registry may say that the institution must be protected because the Internet cannot afford loss of registry continuity. The second part is true. The first part depends on what "protected" means. Protecting continuity should mean protecting records, services, customers and legitimate updates. It should not automatically mean insulating the institution from legal accountability, preserving expansive interpretations of its authority or treating challenges to governance as threats to the Internet itself.
Receivership makes the functional point visible. A court-supervised process can preserve the registry while changing or replacing ordinary leadership. That does not prove receivership is an ideal governance model. It proves that continuity is not identical to board prestige. If continuity can survive emergency governance, then the correct policy question is how to design continuity deliberately: backups, audit trails, succession rules, independent dispute processes, service failover and clear separation between routine maintenance and value-moving enforcement.
IPv6 does not make this architecture unnecessary. During a long transition, more systems depend on both IPv4 and IPv6. That increases rather than decreases the need for clean continuity categories. Which actions are routine? Which require board authority? Which require independent review? Which can proceed during litigation? Which services must remain available even if a resource is disputed? Which records should carry dispute metadata without disabling operations? These questions define real resilience.
The political economy is uncomfortable for registry institutions because it narrows their claim. It says the function is indispensable, but the operator must be replaceable, auditable and constrained. That is the opposite of using IPv6 transition to reinforce registry legitimacy. A registry that truly believes in transition should welcome this discipline. If IPv4 is a declining dependency, then the remaining years of IPv4 dependence should be made as boring, predictable and low-discretion as possible. Institutional drama should be removed from the path, not converted into a proof of institutional importance.
AFRINIC's receivership therefore supplies a better continuity lesson than official reassurance alone. The market does not need symbolic confidence. It needs a registry record that can survive the registry's own governance failures.
Board legitimacy is an economic variable
Board legitimacy can sound like internal governance housekeeping. For a registry, it is an economic variable. A board controls budgets, senior appointments, legal strategy, enforcement posture, policy implementation, risk appetite and the public signals that tell members whether the institution is stable. When the board is absent, contested or elected under a cloud, resource holders cannot treat the problem as ceremonial.
Public reporting in 2025 described AFRINIC as having operated without a board since 2022. The election process under receivership became the subject of legal challenges involving voting rights, nomination procedures and the status of Cloud Innovation in corporate records. Internet Governance Project reported that on 19 June 2025 the Supreme Court of Mauritius dismissed challenges and allowed the election to proceed, while also noting concerns and competing claims around influence over the new board. The Register later reported that voting was suspended and then annulled amid allegations relating to powers of attorney and voter documentation. AFRINIC's path back to ordinary governance was therefore not a clean administrative reset.
For operators, this matters because board legitimacy affects confidence in future registry action. If a transfer is approved, will it be respected? If a policy is implemented, will it survive challenge? If a resource review begins, is it routine or factional? If litigation returns, who decides what the registry can do? If members disagree, which process is authoritative? If the board speaks, does the market accept that it speaks for the institution?
These questions become more important during IPv6 transition, not less. The transition period is full of discretionary choices: how strongly to promote IPv6, how to manage remaining IPv4, how to handle transfers, how to treat leasing, how to interpret out-of-region use, how to prioritize record accuracy, how to fund services, how to structure appeals and how to communicate continuity. A legitimate board can make these choices with more market confidence. A contested board turns each choice into a possible risk event.
Board legitimacy also shapes registry incentives. A leadership team trying to prove institutional strength after crisis may be tempted to use enforcement, rhetoric or policy assertion to restore authority. A leadership team trying to avoid controversy may delay hard decisions. A leadership team dependent on a narrow procedural coalition may favor insiders. A leadership team under legal threat may overcorrect. In each case, the resource holder faces uncertainty not only about policy text but about institutional psychology.
The IPv6 narrative can again obscure the issue. If the official story says the future is abundance and IPv4 is a legacy problem, then board disputes over IPv4 stewardship may seem less important. But the board's control over the current ledger remains economically decisive. A board that cannot command confidence over IPv4 records cannot make IPv4 scarcity disappear by invoking IPv6. It can only transfer risk to operators.
A mature transition policy would make board legitimacy less economically dangerous by limiting discretionary control over scarce records. The more routine, auditable and rule-bound the registry function is, the less each board dispute matters to address value. That is the point of institutional design. Good governance is not a demand that markets trust the right people. It is a system that reduces how much markets must trust any people at all.
Official transition language belongs in evidence, not as the frame
Official materials have evidentiary value. AFRINIC's exhaustion page records Phase 2 facts. Its about page describes its role as a member-based nonprofit under Mauritian corporate law serving Africa and the Indian Ocean region. Its policy manual describes policy development, IPv4, IPv6, transfers, reverse delegation and related services. These materials help identify what the institution says it does and what procedures it publishes.
They should not be allowed to supply the article's frame. That is especially important in a subject like IPv6 transition, where official language tends to merge technical necessity, institutional legitimacy and moral progress. Official transition statements often imply that IPv6 is the natural solution to scarcity, that registries are neutral stewards guiding the community, and that policy process reflects broad legitimacy. Those claims require independent analysis. They cannot be accepted as conclusions simply because they appear in institutional materials.
The same caution applies to global coordination bodies and other registries. Their statements can be useful factual exhibits: who wrote what, when, under which process, and with what asserted authority. But they are not neutral economic proof. Institutions that benefit from the existing registry order have incentives to defend its legitimacy, especially when one registry is in crisis. Their emphasis on continuity may be technically sincere and institutionally self-serving at the same time.
A stronger analysis begins with behavior. What do operators buy? What do customers require? What do courts restrain? What do lenders diligence? What do address markets price? What does litigation reveal about the value of records? What do registry actions do to running networks? What do counterparties accept as evidence? These questions test the transition narrative against real incentives.
On that test, IPv6 looks like a long-term technical necessity but not a medium-term substitute for IPv4 governance. Operators continue to maintain IPv4 compatibility. Transfer and leasing markets continue to exist because demand exists. AFRINIC Phase 2 rules ration what remains. Address-theft allegations and resource disputes show that IPv4 records have real value. Receivership and board disputes show that registry continuity is not a theoretical concern. Customer and counterparty requirements show that IPv4 still prices trust.
Official language becomes dangerous when it reverses this order of evidence. If the institution says IPv6 will ensure growth, and that statement is treated as the frame, then operators asking for IPv4 clarity can be depicted as obstacles. If the institution says it is the steward of the region, and that statement is treated as the frame, then market challenges can be depicted as attacks on the region. If the institution says continuity requires institutional protection, and that statement is treated as the frame, then legal accountability can be depicted as instability. The frame decides who appears reasonable before the facts are weighed.
For a credible policy, official materials should be reduced to their proper role. They can show dates, rules, institutional form, service descriptions and published procedures. The conclusion must come from the political economy of coexistence: IPv6 grows, IPv4 remains scarce, registry records remain powerful, and institutional accountability must therefore become stricter rather than looser.
Gatekeeping borrows the language of transition
Gatekeeping rarely presents itself as gatekeeping. It presents itself as stewardship, fairness, conservation, community, transition or security. Each word can be legitimate in a narrow context. Together they can hide the fact that a private registry retains practical control over scarce public identifiers whose value is created by operators and their customers.
The IPv6 transition strengthens this vocabulary. A registry can say it is not restricting the market; it is guiding the region toward the future. It can say it is not delaying recognition; it is ensuring responsible use during a scarce transition. It can say it is not preserving institutional discretion; it is protecting community-developed policy. It can say it is not resisting assetization; it is preventing speculation while IPv6 adoption proceeds. The words sound technical and public-minded. The effect can be to keep operators dependent on permission.
The AFRINIC controversy around out-of-region use shows the mechanism. Internet Governance Project's 2021 analysis described AFRINIC's dispute with Cloud Innovation as involving claims about regional use, leasing and whether resources allocated in the AFRINIC region could be used outside it. The analysis argued that AFRINIC's attempt to reclaim resources was an overreaction to past problems and that the premise of rigid regional confinement was flawed in a global address market. Others strongly disputed Cloud Innovation's conduct and litigation tactics. The point for transition politics is not to decide every claim in that dispute. It is that the registry's interpretation of acceptable use became economically existential for a resource holder and its customers.
A future-oriented IPv6 frame can make such discretion easier to defend. If IPv4 is treated as a fading legacy resource, then aggressive control over its use appears less consequential. If IPv6 is presented as the proper path for growth, then commercial uses of IPv4 can be cast as obstruction or arbitrage. If transition is treated as a collective project, then individual resource-holder rights can be subordinated to institutional narratives about the region. Scarcity, meanwhile, remains real enough for the controlled resource to be valuable.
This is the contradiction. If IPv4 were truly obsolete, the gatekeeping power would be unimportant. If the gatekeeping power is important, IPv4 is not obsolete. Institutions cannot have it both ways. They cannot minimize IPv4's economic significance when asked for accountability and maximize it when asserting control.
Gatekeeping also survives through procedural complexity. Operators focused on running networks may not attend every policy meeting, track every mailing-list argument or understand every bylaw dispute. Insiders who master procedure can present outcomes as community will even when most affected firms are absent. The transition narrative can add another layer: those who contest the process can be portrayed as resisting progress. The result is legitimacy by exhaustion.
A healthier transition would move in the opposite direction. It would narrow registry discretion as IPv4 becomes more valuable. It would make transfer and lease recognition clearer. It would define limited grounds for intervention. It would protect running services during disputes. It would publish metrics on routine service performance, not only IPv6 adoption. It would give members confidence that the registry is a ledger first and a policy actor only within clear limits.
The long transition should be a reason to reduce gatekeeping power, not to renew it under a different name.
African operators need both adoption and stewardship
Africa's networks do not benefit from a false choice between IPv6 adoption and IPv4 rights. They need both. IPv6 can reduce future dependence, support cleaner network design, improve long-term scalability and align new deployments with a larger address space. IPv4 stewardship protects current customers, contracts, integrations, public services, enterprise systems and the market value of scarce resources during the years in which IPv4 remains necessary.
The false choice is costly because it splits the operator's reality into two political camps. In one camp, IPv6 is treated as progress and IPv4 as hoarding. In the other, IPv4 is treated as capital and IPv6 as vendor or registry propaganda. Real operators occupy a middle position. They deploy what works, conserve what is scarce, buy or lease what they must, satisfy customers, manage vendors, answer regulators and keep services running. Their problem is not ideological purity. It is operational solvency.
AFRINIC-region operators face especially complex coexistence. Some serve lower-income customers with little ability to absorb price increases. Some serve rural or island markets where failure has high social cost. Some are small entrants using upstream addresses while trying to build independence. Some operate mobile networks with huge subscriber scale and mixed device populations. Some run data centres and hosting services that compete against global platforms with deeper address inventories. Some serve public agencies whose procurement files lag behind technical best practice. All of them need a transition policy that respects both future architecture and present constraint.
IPv6 adoption should therefore be judged by practical reduction of future dependency, not by rhetorical conformity. Useful questions include: which customer segments can run IPv6 without service degradation; which public services can publish dual-stack reliably; which enterprise products can reduce dedicated IPv4 demand; which mobile devices and applications actually use IPv6; which procurement clauses can be revised; which security tools support IPv6 evidence; which training reduces operational errors; and which new networks can be designed IPv6-first without sacrificing reachability.
IPv4 stewardship should be judged by different questions: are registry records accurate; are transfers predictable; are lease-related records and routing evidence clear; are disputes isolated; are reverse DNS and RPKI maintained; are contacts correct; are routine changes timely; are enforcement actions proportionate; are appeals credible; are service-continuity categories defined; and are resource holders protected from arbitrary reinterpretation after they have built services on the resources.
Conflating the two sets of questions produces bad policy. High IPv6 deployment does not excuse weak IPv4 records. Strong IPv4 rights do not excuse neglecting IPv6. A region can fail in either direction: clinging to IPv4 without preparing for future growth, or preaching IPv6 while leaving the scarce present vulnerable to institutional risk.
The AFRINIC case points toward a dual mandate. Encourage IPv6 as engineering modernization. Treat IPv4 as economic infrastructure during the coexistence period. The registry's legitimacy should come from doing both competently, not from using one to avoid accountability for the other.
Separation is the core policy principle
The policy lesson is separation. Separate IPv6 adoption from IPv4 stewardship. Separate registry records from institutional authority. Separate routine maintenance from enforcement. Separate dispute notation from service disruption. Separate community rhetoric from member accountability. Separate official factual exhibits from official conclusions. Separate the future state of the Internet from the present rights of operators.
Separation is not bureaucratic neatness. It is a way to reduce moral hazard. When functions are bundled, an institution can use strength in one area to defend weakness in another. It can say that because IPv6 training is useful, its IPv4 discretion should be trusted. It can say that because registry continuity is necessary, its board should be protected from challenge. It can say that because policy is community-developed, specific enforcement is legitimate. It can say that because scarcity is dangerous, market mechanisms are suspect. Separation forces each claim to stand on its own.
For IPv6, separation means adoption metrics should not become a referendum on IPv4 asset legitimacy. A network that leases or sells IPv4 can still deploy IPv6. A network that defends transferability can still support IPv6. A regulator that promotes IPv6 can still require strong due process around IPv4. A registry that allocates IPv6 can still be judged harshly for weak IPv4 record discipline.
For IPv4, separation means scarcity management should not become a license for broad control over operator business models. The registry can require accurate records, prevent fraud, avoid duplicate registration, maintain contact data, publish routing-related information and implement clear policy. It should not convert every change in customer geography, service mix or commercial use into an opportunity for discretionary re-approval unless a specific, lawful and proportionate rule supports that action.
For continuity, separation means the registry's essential services must survive governance failure. Records should be auditable and replicated. RDAP and Whois services should have continuity planning. Reverse DNS and RPKI should have clear maintenance and succession rules. Routine updates should continue where they do not alter disputed economic positions. Courts and independent processes should be able to preserve the last verified operational state while disputes are resolved. The point is to protect the live network, not to protect the symbolism of the incumbent institution.
For AFRINIC, separation would reduce the economic stakes of each institutional conflict. If resource holders know that routine services will continue, that disputed claims will be marked rather than weaponized, that transfers have clear evidence standards, that board changes do not rewrite the ledger and that IPv6 promotion is not a substitute for IPv4 due process, the risk premium falls. Capital can plan. Customers can trust. Operators can deploy IPv6 without fearing that doing so weakens their present IPv4 position.
Separation is also fairer to the registry itself. It lets AFRINIC defend the functions it truly needs to perform without carrying inflated claims about continental authority or moral ownership of the transition. A narrower registry can be more legitimate because its promises are specific enough to test.
Credible transition architecture measures coexistence quality
A credible transition architecture would not measure only IPv6 adoption. It would measure coexistence quality. The practical question is not how loudly a region endorses IPv6, but how safely networks can operate while IPv6 grows and IPv4 remains necessary. That requires metrics that expose registry performance, market friction and customer risk.
For IPv6, the usual metrics still matter: capability, traffic share, prefix allocation, mobile support, enterprise readiness, public-sector deployment, cloud availability and training. But these should be connected to actual reductions in IPv4 dependency. Which services moved without keeping the same IPv4 demand? Which public agencies removed IPv4-only requirements? Which enterprise security tools accepted IPv6 evidence? Which customer segments no longer need public IPv4 exceptions? Which application classes still fail? Adoption without dependency reduction is progress, but it is not retirement.
For IPv4 stewardship, the metrics should be more institutional. Average time for routine contact updates. Time to process reverse DNS changes. Time to issue or update routing-related records. Transfer review timelines. Number of disputed resources and how disputes are marked. Number of service interruptions caused by non-technical governance action. Appeal outcomes. Publication of audit trails for material changes. Clarity of evidence standards. Availability of continuity plans for RDAP, Whois, reverse DNS and RPKI. These measures would tell operators whether the registry is becoming more reliable during scarcity.
Market metrics also matter. How much address demand is met through transfers, leases, upstream assignment, conservation and IPv6 substitution? Where do small operators face the highest friction? Which sectors still require IPv4 compatibility? How often do customers reject IPv6-only service? How do banks, public agencies and data-centre tenants diligence address records? These questions turn transition from ideology into an empirical programme.
AFRINIC's governance history suggests another metric: institutional risk isolation. If a court case, board dispute or election challenge occurs, which services are protected automatically? Which decisions are paused? Which records are frozen in their last verified state? Which changes remain allowed? Which stakeholders receive notice? How is customer impact assessed? A registry that can answer these questions has moved from continuity rhetoric to continuity architecture.
A credible architecture would also state what IPv6 success does not authorize. It does not authorize weakening IPv4 title-like confidence before counterparties have moved. It does not authorize retroactive reinterpretation of old allocations without due process. It does not authorize treating market use as illegitimate merely because scarcity creates profit. It does not authorize vague enforcement over customer geography. It does not authorize using official transition goals to expand registry discretion.
The final measure is boringness. Can an operator present its address plan to a board, bank, public buyer or enterprise customer without spending half the meeting explaining registry politics? Can it say: IPv6 deployment is proceeding; IPv4 compatibility is covered; registry records are clean; continuity risk is bounded; disputes have process; and the remaining scarcity is priced rather than arbitrary? If yes, transition is working. If no, adoption statistics are concealing institutional risk.
The medium term requires discipline, not slogans
The temptation in Internet governance is to end every transition discussion with a slogan. IPv6 is inevitable. IPv4 is legacy. Scarcity is harmful. Markets are exploitative. Community stewardship is neutral. Registry continuity is paramount. Each slogan contains a fragment of something recognizable. None is adequate for AFRINIC's medium-term reality.
The medium term is the hard part. In the long run, more services should be able to rely on IPv6. In the short run, operators must keep existing customers and counterparties working. Between those horizons lies a period of mixed incentives, scarce IPv4, uneven customer readiness, vendor pressure, public-sector conservatism, address markets, registry disputes and legal uncertainty. Political economy lives in that middle period because that is where institutions can use the future to control the present.
AFRINIC is not the only registry facing post-exhaustion questions, but its crisis makes the stakes unusually clear. The region entered Phase 2 scarcity. Reported address misappropriation damaged confidence. A major resource dispute escalated into court action and bank freezes. Receivership separated the registry function from ordinary board control. Election disputes exposed the economic importance of institutional legitimacy. Through all of this, African networks still had to serve customers, deploy IPv6, maintain IPv4 compatibility and satisfy counterparties who care less about governance rhetoric than about service continuity.
The right conclusion is not institutional nihilism. Africa needs reliable number registration. Operators need uniqueness. Customers need stable network identity. Courts need understandable records. Security systems need accurate publication. IPv6 deployment needs coordination. None of that disappears because a registry has failed before or because its authority should be constrained. The question is what kind of authority is justified after scarcity has made the ledger economically valuable.
The answer is disciplined authority. AFRINIC should be strongest where the function is narrow: uniqueness, accuracy, publication, continuity, evidence, fraud control and predictable maintenance. It should be weakest where institutional temptation is highest: discretionary approval of business models, political claims over regional destiny, retroactive reinterpretation, vague enforcement, symbolic continuity arguments and using IPv6 as moral cover for IPv4 control. The registry earns legitimacy by making itself less arbitrary.
For operators, the practical stance is equally disciplined. Deploy IPv6 because future networks need it. Keep IPv4 because customers still price trust through it. Defend transferability because scarce resources must move toward demand. Demand clean records because markets cannot work without them. Resist narratives that treat current assets as shameful merely because a future protocol exists. Do not let a transition plan become a waiver of rights.
For policymakers, the task is to separate adoption from stewardship. Set IPv6 goals, but also require IPv4 ledger accountability. Encourage modernization, but protect running networks. Support registry continuity, but define continuity as service and record preservation rather than institutional immunity. Recognize scarcity, but channel it through transparent markets and due process instead of discretionary gatekeeping. Treat official materials as evidence, not as final judgment.
The opening review meeting then has a better answer. The operator can tell its board that IPv6 is a real programme, not public relations. It can also say that IPv4 scarcity remains a medium-term financial and governance risk. It can assign capital to both. It can ask vendors for IPv6 capability without accepting unnecessary complexity. It can ask the registry for clean IPv4 records without apologizing for scarcity. It can tell customers the truth: the Internet is not moving from one protocol to another overnight; it is moving through a long, uneven coexistence in which trust will attach to whichever identifiers counterparties can still understand.
That is the political economy AFRINIC forces into view. IPv6 deployment does not erase IPv4 scarcity. It does not dissolve registry power. It does not excuse weak ledger discipline. It does not make board legitimacy irrelevant. It does not remove the need for court-resilient continuity. It changes the future path while leaving the present economy governed by installed systems, counterparty trust and scarce records.
The transition will be credible only when institutions stop using IPv6 to avoid the consequences of IPv4's value. Until then, the most important reform is not another slogan about the future. It is a stricter rule for the present: protect the ledger, constrain the gatekeeper, and let operators build the future without surrendering the capital and continuity they still need today.