The diligence file begins with a single IPv4 block. It is not a spectacular hoard. It is large enough to matter to a hosting company, small enough to be traded without making headlines, and valuable enough that nobody in the transaction can treat the registry entry as clerical decoration. The buyer's lawyer asks the first ordinary question: who is the recognised holder? The broker answers with a link to public registration data. The network engineer checks whether the prefix is routed. The abuse desk looks for contacts that still work. The operations team checks reverse DNS. The security team asks who can issue, preserve or withdraw route-origin attestations. The finance team asks whether the seller can deliver clean recognition rather than only a promise in a private contract.
Then the questions become less ordinary. The organisation named in the record may have changed name. The listed contact may be a role account that no longer reaches the right people. The historical allocation may have been made before IPv4 scarcity gave the block a liquid market price. The block may have passed through a lease, a merger, a dormant company, a customer delegation, a disputed authority letter, a court restraint or a review that was never fully resolved. The route may still work while the registration record is stale. The record may look calm while the underlying authority is contested. A false entry may not break traffic today, but it can break settlement tomorrow.
That is why database accuracy is market infrastructure. It is not back-office hygiene. It is the layer on which price discovery, transfer finality, leasing confidence, security reliance, reverse-DNS continuity, abuse handling, lender comfort and member trust are built. In the IPv4 era of abundance, a wrong record was still a problem, but a replacement block might be obtainable and the commercial stakes were lower. In the scarcity era, the same wrong record becomes a risk premium. It changes what a buyer will pay, what a seller can warrant, what a lessee can promise, what a lender will discount, what a court must preserve, and what an operator can safely build on top of a scarce identifier.
AFRINIC is the most revealing case because its institutional crisis turned the registry database from background infrastructure into a visible market surface. The African Network Information Centre is the regional internet registry for Africa and parts of the Indian Ocean region. Its public materials describe services around IPv4, IPv6 and autonomous system numbers, along with WHOIS, RDAP, reverse DNS, an internet routing registry and resource certification. Its policy manual says public IPv4 assignments and allocations must be globally unique and registered in the AFRINIC WHOIS database, and that registration is necessary both for uniqueness and for troubleshooting. It also says registration data must be correct at all times. Those are not ceremonial statements. They are the operating assumptions of a market.
The record has had to carry more weight than its designers could have expected. Public reporting in 2019 described alleged manipulation of address records involving valuable dormant or defunct IPv4 space and companies linked to a former senior AFRINIC staff figure. The Internet Governance Project later treated that episode as part of the background to AFRINIC's aggressive clean-up posture. The Cloud Innovation dispute then raised a different question: when a registry sees a mismatch between recorded use, actual use, regional expectations and commercial leasing, is it correcting the record or trying to control the asset? Litigation, a reported freeze of up to $50m of AFRINIC funds, receivership, board paralysis, election disputes and continuing court fights made the question economically concrete.
The first point is simple and often missed. A registry database is valuable because strangers can rely on it without renegotiating history. A buyer cannot personally verify every allocation letter since the early 2000s. A lessee cannot audit every corporate succession in the holder's past. An upstream cannot investigate every customer chain behind a prefix before accepting a route. A mail operator cannot telephone a continent before judging a reverse-DNS problem. A court cannot reconstruct the full internet-numbering system each time a litigant asks for urgent relief. The registry record condenses those facts into a recognised state. Accuracy is the discipline that makes the condensation tolerable.
That discipline is different from transparency in the broad civic sense and different from the protocol interface by which registration data is exposed. WHOIS and RDAP matter because they publish or deliver registration facts, but the economic question here is not whether one query format is better than another. It is whether the facts behind the query are true, current, attributable, auditable and reliable enough to support commercial reliance. A beautiful interface to stale data is still a market hazard. A public record with false authority can be worse than no public record, because it invites reliance while hiding the defect.
Nor is database accuracy the same as enforcement power. A registry that cannot correct forged authority, stale contacts, lapsed corporate identity, mistaken reverse-DNS delegations, unauthorised route-origin publication or misappropriated address space is not neutral. It is unsafe. But a registry that converts every discrepancy into a discretionary review of a member's business model becomes a different kind of risk. The market needs accuracy, not arbitrary permission. AFRINIC's crisis matters because it shows how quickly the two can become entangled when IPv4 scarcity gives each database field a balance-sheet consequence.
The record that settles the asset
IPv4 addresses are not land, shares or bank deposits. Registries and many official commentators resist property language, and with reason. Address space must remain globally unique, portable in the routing sense, coordinated across regions and attached to operational responsibility. But avoiding property language does not remove economic reliance. A block can support recurring hosting revenue, customer contracts, acquisition value, leasing income, route reputation, firewall allowlists, cloud deployments, anti-abuse systems, reverse-DNS dependencies and continuity plans. The holder may not own the number in a conventional proprietary sense, yet the holder and its counterparties plainly rely on recognition.
The registry database is where recognition becomes legible. It answers a sequence of economic questions. Which organisation is recorded as the holder? Which contacts can act? Which resources are covered? What status is assigned? Are there sub-allocations or assignments that explain downstream use? Is reverse DNS available because the necessary assignments are registered? Can a transfer source be treated as the current recognised holder? Has a block become subject to a dispute? Can the party requesting a change prove authority? If the answers are clear, the block is more liquid. If they are unclear, every counterparty adds a discount.
This is settlement infrastructure in the same practical sense that a land register, securities depository or payment-clearing ledger settles reliance in its own domain. The analogy is imperfect, but it is useful. The value is not that the record creates all economic substance. The value is that the record reduces the cost of proving who may act. If a land registry is frequently wrong, property prices fall in affected files. If a securities depository cannot reliably show entitlement, trading becomes slower and more expensive. If a payments ledger cannot settle balances, merchants raise prices or refuse exposure. If an internet-number registry cannot keep reliable resource records, address markets and network operations price uncertainty.
AFRINIC's own policy language acknowledges this more directly than its institutional rhetoric sometimes does. The manual's goals for IPv4 include uniqueness, registration, aggregation and conservation. Registration is not presented as optional decoration; every public internet address assignment and allocation is to be registered in the database. The manual says unregistered resources are considered invalid and that registration data such as name, block, range, contacts and status must be correct at all times. In reverse DNS, the policy ties delegation to registered assignments or sub-allocations. For ASNs, public registration in the WHOIS database is required and details must be updated when information changes.
Those clauses turn accuracy into an operating condition. If reverse DNS depends on a registered assignment, a stale assignment is not merely an archival flaw. It may impair mail reputation, diagnostics, security tooling and customer support. If transfer source eligibility depends on being the recognised rights holder and not being involved in a dispute, an inaccurate record changes whether a sale can settle. If abuse contacts are published through objects tied to inetnum, inet6num and aut-num records, stale contacts increase response costs and may punish innocent holders through reputation loss. If RPKI authority rests on registry-recognised control, an inaccurate authority chain can affect routing-security reliance.
The market distinguishes between address count and clean address count. A /24 with current contacts, an active holder, no dispute, working reverse DNS, clear abuse responsibility, stable RPKI authority and an obvious transfer path is not the same economic good as a /24 with an old company name, unreachable contacts, uncertain holder authority, a possible historical claim and unclear route-origin control. The numbers may be identical in binary form. The settlement quality is different. Price discovery therefore begins with database diligence.
The same is true for larger blocks, where the price of uncertainty compounds. A /16 can represent tens of thousands of addresses and millions of dollars of implied market value at scarcity-era prices cited by public analysts. If the record behind such a block is ambiguous, the buyer does not merely worry about administrative tickets. It worries about whether closing can occur, whether escrow should hold back payment, whether a future claimant can challenge the chain, whether customer migration should be delayed, whether lenders will accept the asset-supported revenue, and whether the registry may later revisit the transaction. Accuracy has become capital-market plumbing.
The important institutional point is that this reliance does not require a registry to bless every commercial use. A record can say who is recognised without endorsing the price paid. It can identify abuse responsibility without approving a customer's business. It can record a transfer without adopting the buyer's corporate strategy. It can support reverse DNS and RPKI without turning the registry into a commercial regulator. The more precise the record, the less discretionary judgement is needed. Accuracy narrows power.
AFRINIC's challenge is that the same database that should narrow power has sometimes appeared in public controversy as the instrument through which power is exercised. When the registry says a recorded use differs from actual use, it may be identifying a real accuracy problem. When it says that difference may justify termination or reclamation, it moves from accuracy into the economics of control. The boundary is not academic. It determines whether market participants see the database as a settlement layer or a revocable permit book.
How scarcity changed the cost of a wrong entry
IPv4 exhaustion changed the meaning of accuracy because it changed the replacement cost of a mistake. AFRINIC's exhaustion materials say the region entered Soft Landing Phase 1 in March 2017 and Phase 2 on January 13th 2020. In Phase 2, the minimum allocation or assignment is /24 and the maximum is /22. The page explains that organisations in other regions had already been unable to obtain large amounts of fresh IPv4 space and that IPv6 deployment was the long-run answer. At APRICOT 2026, The Register reported an AFRINIC executive as saying the registry still had 773,376 unallocated IPv4 addresses and wanted the remaining pool to reach zero so the conversation could move to IPv6.
The technical aspiration does not eliminate the economic transition. IPv6 may be plentiful, but IPv4 remains needed for broad reachability, customer compatibility and dual-stack operations. Scarce IPv4 therefore becomes an input that must be reused, transferred, leased, cleaned up and documented rather than simply requested from a free pool. In that environment, the value of an address block no longer lies only in its routing utility. It also lies in whether the registry record can support movement from a lower-valued use to a higher-valued use without excessive uncertainty.
A wrong entry in the allocation era was a data-quality defect. A wrong entry in the scarcity era can be a mispriced asset, a blocked transfer, a cloud-capacity delay, a contested lease, a reputational liability or a litigation trigger. If a dormant company's block is incorrectly shown as controlled by someone else, the error can move millions of dollars. If a registered holder is real but the contacts are stale, a valid transfer may be delayed long enough to kill a transaction. If a sub-allocation is not registered, reverse-DNS delegation may become harder and abuse routing may fail. If a block is marked, treated or rumoured as disputed, a cautious buyer may walk away.
Public reporting around AFRINIC's 2019 address-record scandal showed the stakes. KrebsOnSecurity reported allegations that a senior AFRINIC figure, Ernest Byaruhanga, had links to companies involved in selling African IP address blocks and that records for dormant or defunct African businesses had been altered. The report described researcher Ron Guilmette's estimate that the affected addresses exceeded $50m in market value. It also noted that the allegations involved IPv4 space that could be attractive to spammers and marketers because address reputation and availability had become commercially valuable. AFRINIC's chief executive at the time said the organisation was investigating.
The point is not to use one report as a criminal verdict on every named party. The point is to observe the economic mechanism. Once IPv4 has a market price, historical record weakness becomes an attack surface. Dormant companies, acquisitions, inactive contacts, shell entities, forgotten allocations and database change authority all become valuable. A false registry update can perform the economic work of theft even before a packet moves. The block need not be physically seized, because the thing being traded is recognised control. A registry database is therefore not merely a directory. It is a custody file for scarcity.
AFRINIC's clean-up impulse after such allegations was understandable. A registry that discovers or suspects record corruption must repair it. It must review history, check authority, reclaim entries moved through false documents, improve internal access controls, and refer suspected misconduct to appropriate legal channels. It must not allow accuracy to become a polite word for inaction. The market would punish that too, because buyers would assume hidden defects behind apparently clean records.
Yet clean-up has its own risk. If the institution cannot separate fraud repair from open-ended review of lawful commercial use, legitimate holders begin to fear that every valuable block can be reopened. That fear is a cost of capital. A buyer prices the possibility that the registry will later ask whether actual use matches a years-old justification. A lessor prices the possibility that commercial delegation will be recast as abuse of the registration service agreement. A lender prices the possibility that a registry correction will impair customer revenue. Clean-up must therefore be accurate, bounded and reviewable, or it becomes another source of uncertainty.
This is the paradox at AFRINIC's centre. The region needs a more accurate database because scarcity made false records profitable. The region also needs limits on database-review power because scarcity made registry discretion commercially destructive. Both propositions are true. Treating only the first as legitimate produces gatekeeping. Treating only the second as legitimate produces tolerance of stale or false records. A market-grade registry has to hold them together: aggressive against falsity, restrained against lawful reliance.
Scarcity also changes the social politics of database accuracy. Members with few resources may resent large historical holders. Development advocates may prefer regional retention. Commercial holders may emphasise liquidity. Security operators may prioritise abuse contacts and route validation. Courts may focus on contractual rights. These interests meet in the database because the database decides who is recognised. If accuracy is weak, powerful actors can exploit defects. If accuracy is discretionary, powerful actors can seek influence over the registry. The answer is not to choose a side in each political argument. It is to make the record's truth conditions as objective as possible.
AFRINIC's corruption repair problem
Corruption repair is the hardest test of database accuracy because it requires power and distrust at the same time. A registry must distrust suspicious records enough to investigate them. It must also distrust itself enough to prevent an investigation from becoming an excuse for unbounded control. AFRINIC's reported address-record corruption episode created exactly this problem. Public reports described historical changes to WHOIS records, blocks associated with defunct or inactive African entities, alleged shell-company use, and sales into markets far from the original holders. The allegations were grave precisely because the registry record was the market-facing proof of control.
A narrow view of accuracy would treat such a case as a chain-of-authority problem. Who was the original holder? Did that entity still exist? If it dissolved or merged, who inherited the relevant operational rights? Who requested the database change? Was the requester authorised? Were internal credentials misused? Did staff follow documented procedure? Was the resulting holder a legitimate successor, a buyer from a legitimate successor, or a party relying on falsified records? Which routes, reverse-DNS delegations, abuse contacts and RPKI objects were created in reliance on the changed record? Which downstream users were innocent?
Those questions are difficult, but they remain registry questions. They are about truth, authority, chronology, evidence and continuity. They do not require the registry to decide whether IPv4 leasing is morally attractive, whether an address market should exist, whether a holder is too profitable, or whether the region's development goals are better served by one commercial model than another. The more clearly AFRINIC frames corruption repair as authority reconstruction, the easier it becomes for courts and counterparties to trust the result. The more it mixes corruption repair with general hostility to commercialisation, the easier it becomes for affected holders to argue that accuracy has become a policy weapon.
The AFRINIC crisis shows why this separation matters. The Internet Governance Project's 2021 analysis argued that AFRINIC's later posture toward Cloud Innovation was influenced by the earlier corruption scandal and a desire to clean up. According to that account, AFRINIC initially raised concerns about discrepancies between recorded usage descriptions and the countries where Cloud Innovation's resources were actually used, inconsistency between original need and actual utilisation, and regional service obligations. Cloud Innovation contested the interpretation. AFRINIC's March 2021 follow-up reportedly sought detailed information and referred to possible termination and reclamation. The parties then escalated into court.
One can accept that AFRINIC had reason to ask questions without accepting that every possible remedy was proportionate. A discrepancy between record and reality is a proper starting point for an accuracy inquiry. A threat to reclaim millions of IPv4 addresses is not a mere database correction. It changes the market's view of every large AFRINIC holding whose use has evolved since allocation. It also changes how holders respond to future accuracy requests. Instead of treating a request as a chance to clean data, they treat it as the beginning of an enforcement file. Accuracy work becomes adversarial.
That is a loss for the registry. Accurate databases require cooperation. Members must update contacts, register assignments, keep abuse details current, disclose corporate changes, correct reverse-DNS records and report authority changes. If they fear that every update may invite a broad review of business model, region, customer base or historical justification, they will minimise disclosure. Some will keep arrangements at the contract layer while leaving the public record thin. Some will resist voluntary correction until forced. Some will litigate first. The database becomes less accurate because the registry attached too much discretionary risk to being accurate.
Corruption repair also requires a remedy ladder. A clerical correction should not look like a fraud case. A stale contact should not be treated like forged authority. A disputed corporate succession should not be handled like a commercial-use violation. A suspicious historical transfer should be held while evidence is checked, but the hold should be classified, time-bound and reviewable. If fraud is proved, severe remedies may be necessary. If reliance by innocent downstream users exists, continuity planning should accompany correction. The public record should be cleaned without using innocent customers as leverage.
AFRINIC's reverse-DNS policy provides a useful small example of how accuracy can be tied to service without becoming arbitrary. The policy says no reverse delegation of administered or allocated address space is allowed unless an assignment or sub-allocation from the specific allocation is registered appropriately in the database. That is a clear accuracy-service link. It tells the holder what fact must exist for the service. The registry does not need to speculate about the morality of the holder's business. It verifies whether the relevant assignment or sub-allocation is registered.
The same discipline could be applied to higher-value accuracy problems. For each review, AFRINIC should identify the exact fact at issue: holder existence, representative authority, assignment registration, abuse contact, transfer source status, recipient eligibility, court restraint, payment standing, RPKI authority, reverse-DNS entitlement or disputed succession. It should state the effect of non-correction. It should distinguish temporary service preservation from new change approval. It should provide a route for independent review before severe impairment. That is how a registry becomes tougher against corruption while less dangerous to legitimate reliance.
The market will not believe general assurances. It will believe patterns. If AFRINIC corrects false records with documented authority evidence, preserves technical continuity while disputes are resolved, publishes aggregate review categories, and applies severe remedies only to defined misconduct, confidence rises. If accuracy reviews appear selective, retrospective or entangled with commercial ideology, confidence falls. A database becomes trustworthy not because it is never contested, but because the method of resolving contest is predictable.
The Cloud Innovation dispute and the accuracy-control boundary
The Cloud Innovation dispute is often presented as a fight over regional use, commercial leasing, litigation tactics or the future of AFRINIC. For the database-accuracy question it is best understood as a boundary event. It forced the registry and the market to ask whether a mismatch between recorded expectations and actual use should be corrected as a data matter, enforced as a contractual matter, or treated as evidence that the holder's entire resource position was vulnerable. The answer was never merely technical, because millions of IPv4 addresses and substantial revenue expectations were at stake.
According to the Internet Governance Project, Cloud Innovation was a Seychelles-based AFRINIC member that received nearly seven million IPv4 numbers and built a business leasing them, with many customers reportedly outside Africa. AFRINIC's 2020 and 2021 correspondence, as summarised by that analysis, focused on differences between registered usage descriptions and actual countries of use, inconsistency between initial need and later utilisation, and whether services originated within the AFRINIC service region. AFRINIC treated those matters as policy and agreement concerns. Cloud Innovation argued that AFRINIC's position would require ongoing approval of ordinary changes in ISP use and would interfere with market response.
The dispute exposed a difference between two meanings of accuracy. The first is descriptive accuracy: the record should reflect who holds the resources, who operates them, who receives abuse reports, what assignments or sub-allocations exist, what reverse-DNS and RPKI authority applies, and whether any dispute affects changes. The second is normative accuracy: actual use should conform to the purpose, region or need that the registry regards as proper. The first meaning supports settlement. The second can become continuing control unless the policy basis and remedy are sharply limited.
Some normative conditions are unavoidable in allocation from a remaining free pool. AFRINIC's exhaustion procedures require justified need, utilisation thresholds and contractual checks. When a registry still distributes scarce resources at administrative prices, it needs rationing criteria. But market transfers and long-running operational use raise a different reliance problem. A holder's business can change. Customers can change. A cloud or hosting provider can serve users across borders. A lessee can use addresses in a different jurisdiction from the legal holder. If each change reopens the original allocation bargain without clear limits, the database becomes a permanent supervisory device.
This is why the Cloud Innovation correspondence was so economically significant. If AFRINIC had asked only for updated contacts, documented sub-allocations, abuse responsibility, accurate country information, reverse-DNS alignment and route-authority clarity, the request would have looked like database accuracy. If it demanded detailed customer-use justification to decide whether the holder's business model still deserved recognition, it looked more like capital control. The difference is not whether the registry may ever ask difficult questions. It is whether the question is tied to a registry fact or to commercial permission.
The threatened remedy matters as much as the question. Reclaiming millions of addresses, or terminating the registration service relationship behind them, would not simply correct a database field. It would threaten customer contracts, lessor revenue, route-origin reliance, reverse-DNS continuity and address reputation. AFRINIC may argue that severe remedies were available under contract or policy. Cloud Innovation may argue that such remedies were excessive and arbitrary. The courts decide legal merits. The market observes the risk that a database dispute can become an existential asset dispute.
The litigation then created a second-order accuracy problem. When AFRINIC's funds were reportedly frozen up to $50m in 2021, the uncertainty moved from one member's resources to the registry's operating capacity. The registry record remained necessary for all members, but the institution maintaining it became entangled in high-stakes litigation. Receivership in 2023, welcomed by the NRO as a way to preserve status and restore governance, was a continuity response. It did not by itself answer the underlying boundary question: how should the registry distinguish correction of inaccurate resource records from control over evolving commercial use?
By 2026, that boundary had not disappeared. The Register reported AFRINIC accusing Cloud Innovation, Larus and related advocacy campaigns of driving litigation and procedural roadblocks, while Lu Heng, associated with Cloud Innovation and Larus, argued that the registry model concentrates high-consequence power over economically critical number resources without matching liability. The Register also reported a separate dispute over Larus's announcement of a leasing platform and AFRINIC's response that a Mauritian court order did not approve or recognise the asserted structure. Cloud Innovation and Larus disputed AFRINIC's characterisation and said the order did not decide leasing or ownership.
Again, the database-accuracy lesson is not that either side's full public narrative should be adopted. It is that recognition language itself had become market-moving. Did a court order support a continuity structure? Did AFRINIC's register or corporate position imply member rights? Did a published claim misstate judicial approval? Did a correction protect accuracy or chill a business model? These are accuracy questions and market questions at once. They show why registries and holders must be precise. In a scarce-resource market, careless claims about recognition create value or destroy it.
An accuracy-centred approach would require both sides to reduce ambiguity. AFRINIC should not imply that every leasing arrangement is suspect if its concrete concern is false attribution of court approval, inaccurate holder responsibility or evasion of transfer policy. A leasing provider should not imply registry or court recognition beyond what the documents actually support. The record should state what is recognised, what is not recognised, what is disputed, what is under review and what services remain stable. The market can price known limits. It struggles to price rhetorical conflict.
Transfers, leasing and the price of clean recognition
The IPv4 transfer market is where database accuracy becomes visible as settlement. A private purchase agreement does not complete the economic transaction if the registry will not recognise the change. Escrow can move money, lawyers can sign documents, and engineers can prepare route announcements, but the buyer wants a database state that counterparties will accept. The transfer source must be the recognised holder. The resource must not be trapped in a dispute. The recipient must satisfy applicable conditions. Contact, status, reverse-DNS and RPKI arrangements must transition without creating operational gaps.
AFRINIC's policy manual frames intra-region IPv4 transfers as a response to exhaustion. It says transferred resources must come from an existing AFRINIC member account or a legacy resource holder in the service region, and that the source must be the current rights holder recognised by AFRINIC and not involved in a dispute as to the status of those resources. It requires AFRINIC to approve the recipient's need and requires the recipient to be an AFRINIC member subject to current policies and the registration service agreement. It also says transferred IPv4 legacy resources will no longer be regarded as legacy resources.
These rules are not just administrative filters. They affect liquidity. If the source's recognised status is uncertain, the asset is discounted. If a dispute can block settlement, counterparties require conditions. If recipient need approval is unpredictable, buyers hesitate. If legacy status changes on transfer, sellers and buyers price the loss of optionality. If regional transfer limits restrict the buyer universe, exit value changes. A database that accurately records status and a process that applies rules consistently are therefore part of the asset's price.
Leasing is more complicated because registry recognition may not change. A holder can remain the recorded resource member while customers use the addresses under contract. In operational terms this is common across hosting, cloud services, dedicated servers and network outsourcing. In market terms it lets holders monetise underused inventory and lets customers obtain capacity without buying a block. In registry terms it raises accuracy questions. Who is responsible for abuse? Who controls reverse DNS? Who can request RPKI changes? Are downstream assignments registered where policy requires it? Does the arrangement hide a transfer in all but name? What happens when the lease ends?
A registry that ignores leasing may preserve formal simplicity while making the public record less useful. Abuse complaints may go to a holder that has little day-to-day control. Reverse-DNS updates may depend on side agreements. RPKI authority may sit with the lessor while operational pressure sits with the lessee. A customer may believe it has continuity while the registry sees only the holder. On the other hand, a registry that treats leasing as inherently suspect without clear responsibility rules may drive the practice further into private contracts and make the record even less accurate.
The better answer is responsibility accuracy. AFRINIC does not need to publish private lease prices, customer contracts or commercially sensitive deployments. It does need a way to keep operational responsibility legible. The record can identify the resource member, relevant contacts, abuse mailbox, registered assignments or sub-allocations where required, reverse-DNS control, route-authority status, and dispute or hold state. It can allow a holder to remain responsible while recording enough delegated operation to protect users and counterparties. It can distinguish temporary customer assignment from transfer of control.
This approach would also reduce transfer-market risk. If leasing is legible, buyers can see whether a block has active operational dependencies. They can plan customer migration, ROA changes, DNS transitions and abuse-contact updates. If leasing is hidden, a buyer may acquire a block that is economically encumbered by private obligations. If leasing is treated as a violation without a clear rule, a buyer may fear that the seller's past customer arrangements will later contaminate the transfer. Accuracy lowers both risks.
The same logic applies to out-of-region use. Internet routes do not obey neat corporate or continental boundaries. A company registered in an AFRINIC country can have customers elsewhere. An African operator can serve global users. A foreign group can invest through an African entity. A cloud workload can shift location while the legal holder remains in the region. If policy restricts certain uses, the restriction should be explicit, prospective and tied to measurable registry facts. If the registry instead discovers out-of-region use through ad hoc review and treats it as a general accuracy defect, the market prices arbitrary enforcement.
The transfer market also needs error classification. A transfer can fail because a document is missing, the signer lacks authority, the source is in dispute, the recipient fails a policy condition, fees are unpaid, a court order restrains action, the block's history is suspicious, or staff need more information. These are different risks. A buyer can cure a missing document. It cannot easily cure a court fight between others. It can accept a fee adjustment. It may not accept a fraud cloud. Aggregate reporting of delay categories, denial reasons and appeal outcomes would lower the information premium around AFRINIC-administered transfers.
Clean recognition is the product. Not ownership. Not moral approval. Not insulation from all disputes. Clean recognition means that the registry can say, with evidence, who is recorded, what status attaches, what services depend on that status, what policy condition applies and what path exists to change it. The more AFRINIC can sell that reliability through practice, the less its members will need to sell uncertainty through discounts.
RPKI, reverse DNS and WHOIS are accuracy dependants
The public tends to notice registry data when a WHOIS or RDAP query is run. Operators notice it when something breaks. RPKI, reverse DNS, WHOIS, RDAP, IRR objects and abuse contacts are often described as services around the registry. Economically, they are dependants of database accuracy. They take the registry's recognition of resource control and turn it into operational signals used by other networks, security systems, mail systems, customers and counterparties.
RPKI is the sharpest example because route-origin authorisations can influence whether networks accept a route as valid. If the holder record is accurate and the account authority is secure, RPKI strengthens the value of the block. It gives upstreams and peers a cryptographic signal tied to a recognised resource position. If holder authority is disputed or account control is compromised, RPKI can become a continuity risk. Who may create or withdraw a ROA? What happens during a member-status dispute? How are existing objects preserved while authority is checked? Can a registry action unintentionally break reachability?
AFRINIC's crisis has not been primarily an RPKI outage story. That is precisely why the issue should be addressed before it becomes one. When a registry is under litigation, receivership, board transition or member-authority dispute, high-impact security services should be protected by service-state rules. Last-known-good ROAs should not be casually disturbed by ordinary administrative disputes. Emergency locks should be narrow, recorded and reviewable. Authority recovery should require strong proof. Severe enforcement should not use RPKI disruption as leverage except under narrowly defined security or legal necessity.
Reverse DNS is less glamorous but commercially important. Mail delivery, diagnostics, hosting control panels, security logs, customer support and reputation systems can all depend on it. AFRINIC's policy ties reverse delegation to registered assignments or sub-allocations, making database accuracy a precondition of the service. If assignments are stale or missing, reverse DNS can become a symptom of record decay. If a member's standing, representative authority or transfer status is contested, reverse-DNS modifications may become difficult just when customers need stability.
A single inaccurate reverse-DNS dependency can produce real costs. A hosting provider may lose mail deliverability. A customer may see warnings in logs. An abuse complaint may go to the wrong desk. A buyer may delay closing until delegations are clarified. A lessor may face customer complaints because the holder controls the delegation but the lessee bears the operational pain. These are market effects of data quality. They are not solved by telling participants that reverse DNS is a secondary service.
WHOIS and RDAP expose the public record, but their market function depends on the quality and interpretation of fields. The organisation name, status, country, contacts, abuse mailbox, mntner-like authority structures, related route objects and timestamps all become part of diligence. A record that faithfully states disputed status may reduce value but increase trust, because it prevents false finality. A record that silently hides a known authority conflict may raise price temporarily but increase systemic risk. Accuracy sometimes means disclosing uncertainty in a controlled way.
The public-record problem also involves privacy and security, but those should not be used to excuse weak economic signals. A registry need not expose personal documents, private contracts or sensitive customer details to give useful status information. It can publish the resource holder, role contacts, abuse mailbox, status categories, relevant dates, transfer completion state, dispute flags, and service-state consequences. It can provide confidential evidence to an independent reviewer when needed. The market does not require voyeurism. It requires reliable indicators.
IRR objects and routing policy records add another layer. AFRINIC's public materials and general RIR practice recognise that routing registries support route filtering and network coordination. If route objects persist after a transfer, or if stale objects imply authority no longer present, filters and counterparties may rely on bad information. If objects cannot be updated because holder authority is unclear, the market sees operational friction. Again, the record's accuracy is what lets the service do its job.
These services also create blast-radius questions. If a member is late on fees, what happens to existing RPKI, reverse DNS and public data? If a transfer is under review, are existing delegations preserved? If a resource is suspected to have been misappropriated years ago, how are innocent current users protected while authority is reconstructed? If a court order restrains changes, what service state remains stable? A market-grade registry would define these states before crisis. Otherwise every operational service becomes another negotiating chip.
AFRINIC's recovery should therefore treat RPKI, reverse DNS and public registration data as accuracy dependants, not peripheral features. The board and staff should be able to show service-state maps: preserved, update-restricted, transfer-held, disputed, security-locked, court-restrained, revocation-pending or restored. Each state should have a reason, evidence threshold, notice requirement, cure path and review route. Such a map would lower fear. Operators can plan around known states. They cannot plan around institutional improvisation.
This is also where the distinction from a separate discussion of RDAP and WHOIS matters. The interface question asks how public registration records should be exposed, queried and balanced against privacy. The accuracy question asks whether the underlying facts can be trusted as market infrastructure. AFRINIC's problem is not solved by better publication alone. It is solved when the record, the services that depend on it, and the remedies around it form a coherent settlement system.
Governance records are market records
Database accuracy is not limited to resource records. In AFRINIC's crisis, governance files became market records too. Who is a member? Who may vote? Who may hold a power of attorney? Who is a registered member under Mauritian corporate law and who is a resource member under AFRINIC's bylaws? Who may sit on a nomination committee? Who may instruct staff or counsel during receivership? These questions may look like association housekeeping, but in a monopoly registry they affect the market because governance controls the policies and discretion around scarce resources.
The 2025 election episodes made the connection visible. The Register reported that AFRINIC had operated without a board since 2022 and that a court-appointed receiver arranged elections. In June 2025, voting was suspended and then annulled after concerns about powers of attorney and voter documentation. South Africa's Internet Service Providers' Association alleged that representatives of resource holders found votes or powers of attorney recorded on their behalf without valid authority; ICANN also raised concerns and sought explanations. AFRINIC's receiver later organised a fresh election, and a board was elected in September 2025, though public reporting still noted likely court challenges and continuing legal uncertainty.
For an ordinary nonprofit, a messy election is a governance embarrassment. For AFRINIC, it is also a signal about record accuracy. If the institution cannot reliably verify who may vote on behalf of a resource member, how confidently can the market assume it can verify who may transfer a block, update contacts, request reverse DNS, control RPKI or appoint a representative in a dispute? The evidence categories overlap: corporate authority, signatures, proxies, powers of attorney, revocations, role accounts, legal status and audit trails.
This does not mean AFRINIC's staff are incapable of resource verification. It means governance accuracy and resource accuracy share the same institutional muscle. Weakness in one makes counterparties worry about the other. A forged or disputed power of attorney in an election context tells the market that authority documents have value and that attackers or interested parties may try to exploit them. A registry that responds by strengthening pre-validation, member notifications, challenge windows and audit trails strengthens both elections and resource records.
The Cloud Innovation corporate-status controversy adds another example. The Internet Governance Project reported in June 2025 that Cloud Innovation had been quietly added as a registered member in documents lodged with the Mauritian Corporate and Business Registration Department, and that the court later described the classification as erroneous and not attributable to AFRINIC or the receiver in the way challengers feared. The specifics matter to corporate law, but the economic signal is broader. A corporate registry entry, like an internet-number registry entry, can create the appearance of authority. Correcting the error mattered because control of AFRINIC's corporate organs had market consequences.
By 2026, The Register reported that debates over AFRINIC bylaws continued, including claims that resource members' rights under bylaws did not map neatly onto registered-member status under Mauritian law. Whether a proposed legal analysis or bylaw change is the right answer is a separate question. The market lesson is that membership classification is not abstract. If resource members believe they can vote, challenge, elect and constrain the registry, but corporate law recognises a narrower set of formal members for certain purposes, every governance dispute becomes an accuracy dispute about institutional authority.
Governance accuracy also affects policy implementation. A board with contested legitimacy may approve budgets, transfer-policy interpretation, bylaw changes, enforcement posture or litigation strategy. If the board's authority is later challenged, market participants ask whether decisions made in the interim are durable. If the receiver's mandate is uncertain, members ask which actions are preservation and which are policy. If ICANN or the NRO intervenes to explain continuity, critics ask whether official support has become a merits position. Each uncertainty creates a premium around AFRINIC-administered resources.
The answer is not permanent suspicion of every elected director or official statement. Suspicion is costly too. The answer is accurate governance records with the same discipline expected of resource records. Member status should be reconciled between bylaws and applicable company law. Voting authority should be pre-validated and confirmed to the member. Powers of attorney should have limits, timestamps and revocation paths. Election challenges should be time-bound and evidence-based. Board authority should be documented. Conflicts should be disclosed. Receiver-era actions should be classified as preservation, restoration or policy implementation.
A narrow registry benefits from boring governance records. If the market does not have to ask who controls the board, whether the board can instruct staff, whether the staff can process transfers, whether a court may undo an election, or whether resource members can meaningfully constrain discretion, the database becomes more valuable. Governance stability is not valuable as institutional pride. It is valuable because it lowers the cost of relying on the resource record.
The monopoly problem
AFRINIC's database matters more because resource holders cannot easily choose another authoritative recordkeeper. A company with AFRINIC-administered addresses cannot move the same historical relationship to ARIN, RIPE NCC, APNIC or LACNIC because it dislikes AFRINIC's data practices. It may route through foreign upstreams, lease to global customers, incorporate elsewhere or litigate in Mauritius, but the registry record remains tied to the regional system. This is the monopoly feature of regional internet-number administration.
Monopoly does not mean bad faith. It means limited exit. Limited exit changes the economics of error. If a commercial database vendor keeps poor records, customers can switch. If a broker is unreliable, counterparties can use another broker. If an upstream is difficult, a large operator may diversify. But if the authoritative registry record is wrong, stale or contested, the holder cannot shop for an alternative authoritative ledger. It must persuade, wait, litigate or discount. That is why accuracy obligations should be stronger in a registry than in an ordinary association.
Limited exit also changes the economics of fees and legal budgets. Members pay not only for a service desk but for the institution that maintains their recognised resource position. When litigation consumes resources, when accounts are frozen, when receivership delays normal operations, or when legal disputes interrupt allocation and transfer work, members bear costs they did not individually choose. The Register reported AFRINIC's statement in 2026 that litigation and procedural roadblocks were obstructing restoration, community initiatives and member strengthening. Cloud Innovation and related critics argued that legal action was a response to structural registry overreach. Both positions point to the same monopoly fact: disputes at the registry layer impose costs on members who cannot exit the layer.
The monopoly feature also gives accuracy failures a distributional character. Large holders can hire lawyers, lobby, publish arguments, seek injunctions, build alternative commercial structures and absorb delay. Small African operators may depend on a few prefixes, limited staff, thin legal budgets and customer goodwill. A stale contact or delayed reverse-DNS update can be annoying for a large cloud provider and existential for a small hoster. A transfer delay can be a line item for a broker and a missed expansion window for a regional ISP. Accuracy is therefore a competition issue.
A monopoly registry's duty is not to make every member equally powerful. It is to make the record predictable enough that differences in power do not become differences in access to truth. Standard evidence requirements, published status categories, clear service effects, aggregate timing data and independent review reduce the advantage of insiders. If only repeat players know how AFRINIC interprets certain discrepancies, the database becomes a private knowledge market. If the rules are legible, smaller members can rely on the same settlement infrastructure as large ones.
Monopoly also creates temptation for mission expansion. A registry that sees itself as the guardian of a region, the defender of development, the protector of the community or the steward of scarce public resources may believe that its monopoly should be used to shape outcomes. Some of those goals may be worthy. But the database is a dangerous instrument for broad social policy. It can freeze movement, lower value, deter inbound supply, produce shadow leases and raise due-diligence costs. A monopoly registry should be strict about accuracy precisely because it must be restrained about everything else.
AFRINIC's official materials list many useful activities: training, capacity building, DNS support, routing security, policy participation and service improvement. Those activities can help the region. They should not obscure the narrower market reliance at the database core. If the registry record is not accurate, training cannot compensate. If transfer recognition is unpredictable, community language will not restore liquidity. If RPKI and reverse-DNS authority are unclear, development rhetoric will not reassure customers. A monopoly utility earns discretion by proving restraint in the function nobody else can perform.
The monopoly problem is also why outside emergency involvement must be careful. ICANN's 2026 intervention in winding-up proceedings, as reported by The Register, aimed to help the court understand AFRINIC's unique role and to clarify that numbering resources administered through AFRINIC should not be treated as corporate assets available for distribution. That is an important factual and continuity point. But official-system support for preserving the registry function should not become a conclusion that every incumbent action or policy interpretation was correct. Protecting the ledger is different from blessing all uses of ledger power.
Members and markets will judge AFRINIC by whether it reduces the monopoly premium. If a block under AFRINIC administration requires longer diligence, higher warranties, more escrow, more legal memoranda and larger discounts than an equivalent block in a more predictable registry environment, the difference is a tax on the region's operators. It may be invisible in official budgets, but it is real in market prices. Database accuracy is the main tool available to lower that tax.
Accuracy is not a call for secrecy
Some institutions respond to data-quality pressure by publishing less. That would be the wrong lesson for AFRINIC. Accuracy does not require maximal public exposure of private files, but neither does it excuse opacity. The economic value of the database lies in shared reliance. If the registry knows a record is disputed but the public cannot see any status, counterparties are misled. If the registry hides delay categories, members cannot tell whether the process is improving. If the registry treats all evidence as confidential, smaller participants must rely on rumours and personal access.
The right distinction is between private evidence and public status. Private evidence includes corporate documents, identification, contracts, legal advice, customer lists, account-security details and sensitive correspondence. Public status can include the recognised holder, resource range, general status, role contacts, abuse mailbox, reverse-DNS delegation state, RPKI availability, transfer completion state, dispute flag, court restraint category, and relevant dates. A registry can protect privacy while still telling the market whether reliance is clean, pending, restricted or contested.
Status vocabulary matters. Disputed should not mean bad. It should mean a change is contested or a claim requires resolution. Locked should not mean punished unless the lock is a sanction; it may be a security hold. Under review should not be a permanent fog; it should have a reason category and expected next step. Legacy should have operational consequences clearly stated. Transferred should indicate settlement finality. Member not in good standing should distinguish payment, authority, legal and security causes if service consequences differ. The market needs words that reduce ambiguity rather than increase it.
AFRINIC can also publish aggregate accuracy metrics without exposing sensitive member files. How many resource records were corrected? How many contact updates were requested? How many reverse-DNS requests failed because assignments were not registered? How many transfer requests were delayed for source authority, recipient need, dispute status, payment standing or court restraint? How many RPKI authority recovery cases occurred? How many abuse contacts failed validation? How long did different categories take? How many denials were appealed or reversed? These metrics would turn institutional reputation into evidence.
Such reporting would distinguish accuracy from public-relations transparency. A glossy annual statement about improved service is useful only if market participants can see operational behaviour. If transfer timing improves, show distributions. If stale contacts are being cleaned, show counts and cure rates. If historical record reviews are under way, show categories and safeguards. If litigation restricts operations, show which services are affected and which remain protected. If board restoration improves capacity, show ticket queues, service states and correction throughput. Accuracy earns trust through boring data.
Privacy objections are real but manageable. Abuse contacts should not expose personal vulnerability where role addresses can work. Fraud reviews should not publish investigative detail before facts are established. Customer lists should not be made public merely because a registry wants to verify use. Court orders may limit disclosure. But privacy should not be used as a blanket for institutional discretion. A registry can redact the evidence while publishing the category. It can protect individuals while revealing method. It can avoid prejudging a dispute while warning that settlement is not final.
The same principle applies to official statements. During AFRINIC's crisis, communiques, letters, court summaries and media reports often moved market expectations. A statement that a court did not approve a leasing structure may be necessary to correct an overclaim. It should be narrow enough not to imply a final ruling on every commercial issue unless that is what the court decided. A statement that the registry is back on track may be encouraging. It should not hide active constraints on allocation, transfers, member files or litigation. Accuracy in communication is part of database accuracy because words change reliance.
Accuracy also requires historical humility. Some old records will be incomplete because early internet-number administration did not anticipate today's market. Some organisations will have changed form multiple times. Some contacts will be gone. Some assignments will have been made under assumptions no longer held. A registry should not pretend the past was cleaner than it was. Instead it should define evidence paths for regularisation: corporate succession, operational continuity, historical routing, payment history, public documents, old correspondence, third-party attestations and court recognition where necessary.
The market can handle uncertainty if it is labelled. It cannot handle hidden uncertainty that later appears as a challenge. A record marked as regularised after documented succession may be more trustworthy than a record that looks old and pristine but has never been checked. A transfer delayed for a published source-authority reason is less damaging than a transfer delayed without explanation. A block under court restraint can still be valued if the restraint's scope is clear. Public status is how private uncertainty becomes priceable rather than poisonous.
What a market-grade accuracy regime would require
AFRINIC's accuracy problem cannot be solved by one clean-up campaign. It needs an accuracy regime. The first component is a resource-record assurance programme. Each high-value or high-risk record should have a confidence state based on holder identity, corporate existence, representative authority, contact reachability, assignment registration, reverse-DNS dependency, RPKI authority, transfer history, dispute status and evidence age. The confidence state need not expose private documents. It should tell authorised parties and, where appropriate, the public whether the record is routine, needs update, under verification, disputed, regularised or subject to legal restraint.
The second component is authority verification that is proportionate to consequence. A password reset, contact update, reverse-DNS change, RPKI authority recovery, transfer request and vote in a board election should not use identical evidence standards, but they should draw from the same authority framework. The more valuable or irreversible the action, the stronger the proof. Members should receive notices when a representative is added, a power of attorney is lodged, a material transfer is requested, or a high-impact service state changes. Silent authority changes are unacceptable in a scarcity environment.
The third component is a remedy ladder. Accuracy defects should lead first to correction, then targeted verification, then temporary hold where necessary, then independent review for disputed facts, and only then severe remedies such as transfer refusal, service restriction, termination or reclamation. Fraud and security emergencies may justify faster action, but the category should be recorded and reviewed after the fact. A stale contact should not be treated like misappropriation. A payment delay should not be treated like forged authority. A customer-use question should not be treated like duplicate resource control.
The fourth component is service continuity by default. Existing routing-security and reverse-DNS states should be preserved where law, security and fraud prevention permit while authority disputes are resolved. New changes may be restricted without breaking last-known-good operation. Customer-impact analysis should precede severe action. If a holder's record is defective, the registry should ask how to correct the record while limiting harm to innocent downstream users. Continuity is not indulgence. It is the economic reason registries exist.
The fifth component is transfer and leasing legibility. Transfers should have published timing targets, evidence categories, denial reasons and appeal routes. Leasing or operational delegation should have responsibility records: holder, operator where disclosed, abuse contact, reverse-DNS manager, RPKI authority, end-of-term cleanup and dispute process. AFRINIC need not endorse every lease to require responsibility clarity. If a lease is being used to evade a transfer rule, the registry should identify the facts that make it an evasion. If it is ordinary delegated use, the registry should make the responsibility chain findable.
The sixth component is historical regularisation. AFRINIC should expect old files to contain defects and should make repair cheaper than concealment. A holder with stale contacts or old corporate documents should be encouraged to regularise without fearing that voluntary disclosure automatically opens a broad commercial review. The programme should distinguish benign historical drift from suspicious record change. It should include clear evidence alternatives for defunct entities, public-sector successors, mergers, universities, research networks and long-running operators whose early records predate modern expectations.
The seventh component is independent review for severe accuracy disputes. When a decision could impair significant resources, prevent a transfer, alter RPKI authority, remove reverse DNS, terminate a member relationship or reclaim a block, the affected party should have access to review outside the ordinary staff line and insulated from board factional pressure. The reviewer need not replace Mauritian courts. It should create a technical and evidentiary record that courts can understand, members can trust and markets can price.
The eighth component is aggregate reporting. AFRINIC should publish periodic statistics on data-quality corrections, contact validation, abuse-contact reachability, reverse-DNS dependency failures, RPKI authority recoveries, transfer timing, denial categories, dispute holds, appeal outcomes, court restraints, member-authority challenges and service-state changes. The reports should be designed for risk pricing, not self-praise. They should show where accuracy is improving and where uncertainty remains.
The ninth component is governance-record alignment. Resource-member status, registered-member status, voting authority, proxies, powers of attorney, committee appointments, board minutes and receiver-era actions should be reconciled with applicable law and bylaws. Governance files should be treated as high-impact records because they determine who can control the monopoly database. Election integrity is not separate from database integrity. It is the same authority problem in another form.
The tenth component is liability alignment. A registry cannot hold high-consequence discretion while treating every commercial loss as someone else's problem. That does not mean AFRINIC should insure every member against market risk. It means severe actions should be constrained by evidence, procedure and review, and the registry should accept that mistakes have consequences. If liability remains minimal, discretion must be narrower. If discretion remains broad, accountability must be stronger. The current model becomes unstable when valuable resources meet disclaimers and open-ended power.
These components are not radical. They describe what a settlement utility should do once the assets recorded in its database become scarce, valuable and contested. They would make AFRINIC more capable against fraud, not weaker. They would give staff clearer authority, courts clearer records, members clearer duties and markets clearer prices. The cost of building them is real. The cost of not building them is hidden in discounts, litigation, delayed transfers, private workarounds and lost trust.
The measure of recovery
AFRINIC's institutional recovery will be easy to overstate. A board can be elected. A budget can be approved. A strategy can be announced. A court can preserve the organisation. ICANN can intervene to explain the registry's systemic role. The NRO can welcome receivership as a step toward restoring governance. Staff can keep services running through years of pressure. All of that matters. None of it proves that database accuracy has become reliable market infrastructure.
The measure of recovery is whether the accuracy premium falls. Do buyers of AFRINIC-administered IPv4 space require fewer special warranties? Do transfer files close with more predictable timing? Do lessors and lessees know how operational responsibility should be recorded? Do reverse-DNS and RPKI changes remain stable during ordinary disputes? Do stale contacts get corrected without provoking unnecessary fear? Do historical records become regularised rather than exploited? Do courts see precise status categories rather than broad institutional claims? Do members trust that authority documents cannot be used silently against them?
Another measure is whether severe disputes stop contaminating the whole registry. A single member dispute should not threaten bank continuity, allocation work, transfer processing, RPKI confidence, election legitimacy and the reputation of all AFRINIC records. A suspected false record should not require a political war. A resource holder's commercial model should not be judged through vague accuracy language. A court order should not become a market myth. A public communique should correct only what it can prove. A registry that can contain disputes has become more accurate in the institutional sense.
AFRINIC will also need to prove that corruption repair and market reliance can coexist. The 2019 reported address-record scandal made laxity unacceptable. The Cloud Innovation litigation made unchecked discretion unacceptable. Those lessons are often treated as belonging to opposing camps. They should be read together. False records must be found and corrected. Legitimate reliance must not be destroyed through retrospective, discretionary or poorly classified action. The registry's job is to maintain a true settlement record, not to choose between fraud tolerance and asset control.
The African context makes the stakes larger, not smaller. Many operators in the region face high capital costs, currency pressure, infrastructure gaps, uneven access to legal advice, dependence on scarce IPv4 and limited ability to diversify registry risk. A database discount imposed by uncertainty is therefore a development cost. It raises the price of network expansion, lowers the value of legitimate holdings, discourages formal transfers, pushes leasing into opacity and rewards actors with procedural sophistication. Clean records are not a luxury for rich markets. They are a way for smaller operators to transact without being punished for lack of insider access.
Official materials can supply factual exhibits but not the conclusion. AFRINIC's mission statements, NRO support, ICANN interventions and policy-process language explain how the system describes itself. The market asks a different question: can the record be trusted under stress? That question cannot be answered by saying the registry is important. It can only be answered by showing that the database is accurate, corrections are bounded, disputes are labelled, services are protected, authority is verified and severe remedies are rare, proportionate and reviewable.
The due-diligence file for the single IPv4 block should eventually become boring. The buyer should be able to confirm the holder, authority, status, reverse-DNS dependency, RPKI control, transfer path and dispute state without commissioning a history of AFRINIC's crisis. The seller should be able to warrant the record without pricing institutional surprise. The lessee should know who carries operational responsibility. The bank should understand the registry risk without excluding the address-supported revenue from its model. The court should see a narrow record rather than a constitutional drama.
That boredom is the highest form of registry success. It means the scarce identifier can move, support customers, secure routes, handle abuse, maintain reverse DNS and carry investment without every counterparty asking whether the record itself may be wrong. AFRINIC's crisis showed how costly non-boring records can become. Its recovery will be real only when accuracy once again makes the registry almost invisible.
The economics are therefore unforgiving. IPv4 scarcity turned AFRINIC's database into settlement infrastructure. Record corruption showed that false entries can capture value. The Cloud Innovation dispute showed that accuracy review can shade into asset control. Receivership and elections showed that governance records affect market trust. RPKI and reverse DNS show that operational services depend on correct authority. Transfers and leasing show that markets need clean recognition more than rhetoric. A monopoly registry that gets these facts right lowers the cost of connectivity. A monopoly registry that gets them wrong taxes every block it records.
AFRINIC's most important asset is not the remaining free pool, nor its official recognition, nor the language of community stewardship. It is the reliability of the database that lets everyone else treat scarce number resources as usable, transferable and supportable inputs into real networks. Accuracy is not a virtue beside the market. It is the market's settlement layer.