Begin with the bill, not the hearing. A law firm's invoice reaches a regional internet registry before any judge has decided who is right. It records hours spent drafting affidavits, answering urgent applications, reviewing membership files, advising on bank orders, parsing election rules, preparing a public statement or weighing the cost of an appeal. In an ordinary company, such a document is an expense claim against uncertainty. In a monopoly registry, it is also a governance signal. It tells members, litigants and counterparties what the institution is willing to defend, what it is willing to delay, what it treats as existential, and how much shared money it will convert into legal endurance before the merits are reached.
AFRINIC makes that signal unusually visible. The African Network Information Centre is the regional internet registry for Africa and parts of the Indian Ocean. It does not run every network in the region and it does not route packets for its members. Its power sits in a thinner and more valuable layer: records. It allocates and registers IPv4, IPv6 and autonomous system numbers; maintains authority information; supports reverse DNS, WHOIS, RDAP, IRR and RPKI functions; and provides the recognised administrative ledger that network operators, customers, brokers, courts and other registries consult when asking who may rely on which number resources. When a registry like that pays counsel, it is spending around the identity layer of the internet.
The first economic question is not whether AFRINIC should have lawyers. Of course it should. A registry without legal capacity would be an inviting target for fraud, abusive claims, injunctions, vendor disputes, employment risk, election manipulation and strategic litigation. The harder question is what legal capacity does to incentives. A budget for counsel can protect the ledger against attack. It can also buy time for an institution to resist accountability. It can preserve member services. It can also turn member fees into fuel for discretionary fights over the scope of registry power. It can give management confidence to take necessary action. It can also let management underprice the downside of severe threats because the cost will be spread across members who cannot take their records elsewhere.
That ambiguity sits at the heart of AFRINIC's crisis. Public reporting has described a sequence few institutions could survive without litigation becoming a central operating function: alleged address-record corruption involving valuable IPv4 blocks and a former senior staff figure; a confrontation with Cloud Innovation over resource use, out-of-region customers and the possible withdrawal of millions of IPv4 addresses; a provisional court order reported to have frozen up to $50m in AFRINIC funds; years without a functioning board; receivership under Mauritian court supervision; election processes strained by claims about powers of attorney and voting authority; later board restoration under continuing legal challenge; a winding-up application; disputes over transfer policy and leasing representations; and repeated interventions or statements by bodies worried about registry continuity.
Those episodes can be argued from several angles. One inquiry might ask who should govern AFRINIC. Another might ask who ultimately bears registry fees. A third might ask how large reserves should be and how they should be protected. The narrower question here is different: how does the ability to spend on lawyers change behaviour before any court decides the merits? The answer is uncomfortable. In a normal market, legal spending is disciplined partly by exit. If a supplier fights customers too aggressively, customers can leave. AFRINIC's members cannot move an African registry relationship to ARIN, RIPE NCC, APNIC or LACNIC because they dislike the legal posture in Mauritius. The legal budget is therefore funded from a captive relationship and spent in a dispute environment whose effects spread beyond the parties.
Every dollar spent on counsel in that environment has two possible meanings. It may be the cost of defending a shared settlement ledger from private capture, fraud or destabilising injunctions. It may also be the price of extending institutional discretion after trust has thinned. Members cannot tell the difference merely by reading the invoice total. They need to know the purpose, authority, cost band, service risk, settlement logic and alternatives. Without that classification, legal spending becomes the least transparent policy process in the registry: not a public debate over resource rules, not a board election, not a fee schedule, but a series of tactical choices that can shift the economic boundary between the registry and its members.
This is not a defence of every party that sues a registry. Large resource holders can use legal budgets strategically as well. A well-funded member can impose costs on a registry, freeze its attention, delay governance recovery, create bargaining leverage and make every unrelated member live with service-continuity risk. AFRINIC's history shows both sides of the hazard. The registry can overreach because it believes it is defending the community; a litigant can overreach because legal pressure can make an essential institution bend. Smaller members, universities, data centres, access networks, public-sector operators and ordinary resource holders then discover that other people's legal stamina has become a hidden risk premium on their own records.
That is why the invoice is the right opening scene. It arrives before the judgment. It prices the path to the judgment. In a monopoly registry under IPv4 scarcity, that path can change the institution more than the final order does.
Litigation as an incentive system
Legal budgets are often discussed as an after-effect of conflict. A dispute arises; lawyers are hired; costs accumulate. AFRINIC shows why the sequence is more circular. The expected budget affects the decision to start, escalate, resist or settle a dispute. If an institution believes it can fund years of proceedings, it may choose a sharper enforcement posture than it would choose under immediate budget discipline. If a member believes it can outlast the institution, or impose enough cost to force relief, it may litigate beyond the narrow issue in dispute. The legal budget is not just an accounting line. It is part of the strategic game.
The registry's side of the game is easy to understand. It has duties that cannot be performed by silence. If it suspects that records have been manipulated, that authority documents are false, that resources are being used outside agreed terms, that a court order threatens registry continuity, or that an election process has been compromised, doing nothing is itself a choice with costs. A thin registry is not a passive registry. It still needs counsel to interpret contracts, respond to orders, verify corporate obligations, preserve evidence, protect staff and keep essential publication services from becoming leverage in a private fight.
But legal capacity changes risk appetite. A lawyer can make a severe step feel manageable by describing the arguments available after the step is challenged. A board or management team can convince itself that a contested action is prudent because the institution has cash, public support, emergency rhetoric or outside sympathy sufficient to withstand the challenge. The cost of that confidence is not borne only by the officials who approve the action. It is borne by the member-funded organisation and, indirectly, by members who must live with the registry's degraded credibility if the action becomes a multi-year crisis.
AFRINIC's confrontation with Cloud Innovation illustrates the mechanism. According to the Internet Governance Project's 2021 account, AFRINIC wrote to Cloud Innovation in 2020 raising concerns about discrepancies between registered descriptions of use and countries where resources were being used, about consistency between justified need and actual purpose, and about whether services originated in the AFRINIC region. In 2021, the registry renewed its position and, according to that account, asserted that it could determine whether to terminate Cloud Innovation's registration service agreement and reclaim IPv4 number resources. Cloud Innovation contested the allegations and treated the threatened remedy as existential.
The merits of that dispute belong to courts and the relevant contracts. The incentive lesson appears before the merits. A threat to reclaim millions of IPv4 addresses is not an ordinary compliance letter. It affects customers, leases, routing reputation, balance-sheet assumptions and future transactions. A registry contemplating such a remedy should expect litigation. That expectation should force a budget question at the decision point: is the remedy so necessary to ledger integrity that the institution should spend member money defending it, or is a narrower remedy available that would protect records at lower systemic cost? If that question is not asked before the letter goes out, the legal budget becomes retrospective justification.
The same mechanism operates on the member side. A large holder facing resource withdrawal may decide that ordinary negotiation is insufficient. It can seek injunctions, damages, disclosure, corporate-law remedies, pressure on bank accounts, challenges to board authority and public arguments about structural risk. Some of that may be lawful self-defence. Some may be strategic escalation. Either way, legal capacity lets the member shift the conflict from a resource file to the registry's institutional survival. Once that happens, the dispute no longer belongs only to the two parties. The legal budget of one member can impose a continuity risk on all members.
This is why "who is right" is not the only governance question. Institutions need rules for the period before rightness is decided. A registry's legal budget should make necessary defence possible without making maximalism cheap. A member's access to court should make accountability possible without making paralysis a bargaining tactic. AFRINIC's experience is a study in what happens when that boundary is underdeveloped: each side can plausibly claim to be defending continuity, and each side can impose costs larger than the immediate legal question.
The institutional economics are not subtle. If the registry is under-capitalised, it may be bullied or captured. If it is over-insulated by reserves, sympathy or opaque legal budgets, it may fight too long and concede too little. If large members can externalise legal disruption onto smaller members, they may litigate too broadly. If courts can freeze operating funds before the merits are examined, provisional relief becomes a systemic instrument rather than a narrow protection. The legal budget is the transmission belt among all four problems.
The monopoly ledger changes the meaning of a bill
AFRINIC's members pay for a service relationship that cannot be substituted in the ordinary way. That fact changes the moral and economic character of legal spending. If a cloud provider, transit seller or software vendor becomes too litigious, customers can attempt to migrate. Migration may be expensive, but it is conceptually available. A resource holder in the AFRINIC region cannot move the authoritative history and administration of the same resources to another regional registry merely because AFRINIC's legal budget has become alarming. The member is bound to a ledger relationship.
That ledger relationship is narrower than sovereignty but stronger than a normal vendor contract. AFRINIC's official materials describe a nonprofit, member-based organisation registered in Mauritius, serving Africa and the Indian Ocean region, distributing and managing internet number resources, and offering registry services including reverse DNS, WHOIS, RDAP, routing registry functions and RPKI. Those facts are useful exhibits. They do not settle the incentive problem. The crucial point is that AFRINIC's record is a recognised coordination layer. It helps others decide whether a block is legitimate, whether a route-origin assertion has a credible basis, whether a reverse-DNS delegation should be accepted, whether a contact is authorised, and whether a transfer or assignment is documented.
When legal spending surrounds such a ledger, the bill is partly a tax on dependency. Members fund the institution because the ledger must continue. They do not all share the same position on a lawsuit, a transfer rule, a leasing dispute or an election challenge. Yet the legal budget is financed from an institutional base to which they are tied. That does not make the spending illegitimate. It makes classification essential. Members need to know whether the bill protects services they cannot replace or finances an institutional theory they might reject if given a direct vote.
The dependency is technical as well as commercial. RPKI publication, reverse DNS, WHOIS, RDAP, IRR entries, account standing, ticket processing and allocation records are mundane until they are uncertain. A network can keep routers running while a registry is in court, but its risk surface changes. Customers ask whether address use is secure. Banks and acquirers ask whether a block can be transferred or encumbered. Security teams ask whether contact records are current. Counterparties ask whether a court order may affect recognition. Lawyers ask whether a registry officer had authority to act. The technical layer and the legal layer do not merge, but the second begins to price the first.
Counsel's advice can therefore function like hidden policy. Suppose a registry's lawyers recommend a litigation position that treats leasing as inconsistent with the registry's view of regional purpose. Suppose they recommend resisting a transfer until a broader policy question is resolved. Suppose they recommend public statements denying that a court order recognised a leasing structure. These may be correct legal positions. They may also affect the price, liquidity and confidence of IPv4 transactions. If such positions are adopted through legal tactics rather than clear member-visible policy, the legal budget has become a channel for economic governance.
The monopoly ledger therefore requires a higher standard than "the lawyers advised it." Lawyers advise within a mandate. Who defines that mandate matters. If the mandate is "win the case", the budget will search for advantage. If the mandate is "protect the narrow ledger, preserve continuity and minimise member-wide externalities", the budget is disciplined by institutional economics. AFRINIC's crisis shows how easily those mandates blur when the institution is under attack and when critics of the institution also have commercial interests.
Members should not demand publication of every legal memo. Privilege exists for a reason. But privilege should not hide the economic category of spending. The registry should be able to say, without waiving confidential advice, that a legal bill concerns bank-access defence, receiver compliance, election verification, resource-status litigation, transfer-policy implementation, winding-up response, bylaw review, employment matters or communications restraint. Each category has a different claim on captive member funding. A monopoly ledger cannot ask members to pay for legal spending while refusing to describe the function the spending serves.
Scarcity made legal posture valuable
IPv4 scarcity is the reason AFRINIC's legal budget matters beyond administrative governance. In a world of abundant, easily replaceable IPv4 supply, a contested registry record would still be inconvenient. In the actual world, IPv4 recognition carries economic weight. Addresses are used in hosting, access, cloud services, customer contracts, leasing arrangements, acquisitions, financing assumptions and continuity planning. Scarcity gives the record a market shadow. Legal action around the record therefore changes value before technical routing changes.
AFRINIC's own public exhaustion materials describe the region's move through soft-landing phases: Phase 1 began in March 2017, and Phase 2 was reached in January 2020. In Phase 2, public materials describe a minimum allocation or assignment size of /24 and a maximum of /22. The Register later reported that in early 2026 AFRINIC still had 773,376 unallocated IPv4 addresses, while an AFRINIC executive expressed a desire to reach zero so discussion could move more fully to IPv6. Whatever one thinks of that aspiration, the economics of the transition are clear. While networks still need IPv4, the remaining pool is small, the transfer and leasing environment matters more, and existing records carry more financial significance.
Scarcity turns legal positions into price signals. If a registry appears willing to challenge out-of-region use, counterparties discount resources that might attract review. If a court protects a holder from reclamation, customers treat that protection as evidence of continuity, even if the final merits remain unsettled. If a receiver or board process is contested, buyers ask whether transfer approvals will later be questioned. If public statements suggest that a leasing structure has or lacks judicial recognition, the market reacts because recognition itself is part of the product being sold.
The 2026 dispute over statements by Larus and Cloud Innovation about a leasing platform shows how precise the issue has become. The Register reported that Larus announced a platform tied to a claimed court-ordered continuity structure, that AFRINIC publicly rejected any suggestion that a court order had approved or recognised such a structure, and that a Mauritian interim order later required restraint against representations falsely attributing judicial approval or endorsement of leasing, monetisation, transfer or commercial exploitation of AFRINIC-allocated resources. Cloud Innovation and Larus disputed AFRINIC's characterisation and emphasised that the order was not a final judgment on leasing or ownership.
The most important fact for present purposes is not which press statement was more persuasive. It is that the parties fought over what the market could infer from a court order. That is a legal-budget incentive in concentrated form. Counsel is paid not only to win damages or avoid liability, but to control the recognition environment in which customers and counterparties decide whether a resource position is safe. Litigation becomes part of market microstructure. A phrase in a court order, communique or takedown application can shift perceived value.
That dynamic makes overstatement dangerous. A registry that overstates the effect of a court order may use legal uncertainty to chill legitimate transactions. A holder that overstates the effect of a court order may market private commercial activity as publicly validated. A court that grants broad interim relief may change bargaining power before a full record exists. A reporter, advocate or trade association that simplifies the dispute may unintentionally move prices. In a scarcity market, language is not cheap.
Legal budgets reinforce the problem because they determine who can keep language contested. A small holder cannot afford repeated urgent applications over wording, platform claims, transfer statements or member authority. A large holder can. A registry with reserves or outside support can respond. Smaller members watch from the side, aware that the resulting precedent may govern their own holdings. The legal contest becomes a club good for well-funded actors and a public risk for everyone else.
AFRINIC's lesson is that scarcity requires legal minimalism. The registry should litigate when the ledger's integrity or continuity is threatened, but it should avoid turning every commercial practice into a test of institutional authority. Holders should defend their reliance interests, but they should avoid turning every ambiguity into a claim that the registry's role has been judicially transformed. Courts should protect parties from irreversible harm, but should also recognise that provisional orders can spill over into a whole region's registry confidence. The more valuable IPv4 becomes, the more carefully legal budgets must be tied to narrow purposes.
Enforcement before the merits
The legal-budget incentive is most dangerous when enforcement is severe and the merits are unsettled. A registry must have enforcement tools. It must be able to correct inaccurate records, respond to fraud, require members to maintain contact information, enforce payment obligations, apply community-developed policies, comply with court orders and protect the uniqueness of number resources. A registry that cannot enforce anything is not neutral; it is unreliable. The problem is not enforcement. It is the combination of severe remedy, disputed authority and member-funded legal capacity.
AFRINIC's post-2019 environment made enforcement pressure understandable. KrebsOnSecurity reported in December 2019 that researcher Ron Guilmette and South African journalists had traced allegations involving dormant or defunct African IPv4 resources, companies linked to Ernest Byaruhanga, a former AFRINIC policy coordinator, and address space estimated by Guilmette to exceed $50m in market value. AFRINIC's then chief executive said the organisation was aware of the allegations and investigating. Byaruhanga had resigned shortly before the reporting. The details were reported as allegations, but the institutional effect was plain: the registry's record-integrity controls were under scrutiny.
After such a scandal, a registry has incentives to prove seriousness. Staff become more cautious. Management wants to show that past weakness is over. Community participants demand cleanup. Outside critics ask whether the registry is competent. Counsel may advise that failure to act increases liability. These are real pressures. They can also produce overcorrection. Fraud repair and policy enforcement are not the same thing. Fraud repair concerns false documents, insider misuse, dormant-company manipulation, fabricated need or stolen authority. Policy enforcement may concern changes in deployment, customer geography, leasing arrangements or the interpretation of terms written before scarcity made every phrase valuable.
The Cloud Innovation dispute sat near that boundary. AFRINIC had reason to care about whether justified-need representations and regional service commitments meant anything. Cloud Innovation had reason to argue that ordinary network use changes, that out-of-region customers are not necessarily fraud, and that withdrawal of existing resources would be disproportionate if the dispute was about interpretation rather than deception. A legal budget can help a registry test those arguments. It can also tempt the registry to lead with a remedy so grave that a court fight becomes inevitable.
The incentive should be reversed. The more severe the remedy, the stricter the pre-litigation budget test should be. Before threatening resource withdrawal, a registry should ask whether a narrower step would protect the ledger: a documentation request, a prospective compliance plan, a dispute flag, an independent review, a transfer hold with reasons, a cure period or a targeted correction. Severe action may still be justified in fraud, clear contractual breach or risk to uniqueness. But the institution should not discover proportionality only after counsel is defending the maximum remedy in court.
Legal budgets can obscure this because litigation reframes the question. Once the case is filed, the registry asks how to defend the action already taken. Counsel build arguments around authority, contract text, policy purpose, procedural regularity and harm. The member builds arguments around reliance, disproportion, selective enforcement, property-like value, customer impact and registry overreach. The question that might have avoided the crisis becomes harder to ask: was there a lower-cost institutional path that protected accuracy without threatening systemic disruption?
This is not an argument for timidity. A registry that discovers genuine record theft or forged authority should preserve evidence, notify affected parties, seek orders where necessary and repair the ledger. But the more it asks captive members to fund that legal effort, the more it must separate fraud categories from economic-control categories. Otherwise the legal budget becomes a way to make past control failures justify future discretion.
AFRINIC's history should make that distinction central. The address-record scandal showed why weak controls are costly. The Cloud Innovation litigation showed why aggressive controls are also costly. Legal-budget discipline is the art of not using one lesson to erase the other.
When a bank freeze prices the whole registry
The reported freezing of AFRINIC's bank accounts in 2021 was a legal event with budget consequences far beyond the parties. According to the Internet Governance Project, the Supreme Court of Mauritius provisionally froze up to $50m in AFRINIC funds held at Mauritian banks after Cloud Innovation claimed large reputational damages in the dispute. The same account questioned the proportionality of freezing operational funds before the underlying claims had been fully tested. Cloud Innovation's side saw legal pressure as protection against existential registry action. AFRINIC's side saw the result as a threat to ordinary operations. Both descriptions show the same institutional problem: provisional relief can convert a bilateral conflict into a systemic liquidity shock.
For a registry, cash is not only corporate comfort. It pays staff, technical vendors, security operations, member support, legal compliance, elections, communications and continuity. A bank freeze therefore changes bargaining incentives immediately. The institution may become more willing to settle because operations are at risk. It may become more defiant because it sees the litigant as threatening the whole community. Members may rally behind the registry because services are endangered. Members may also ask why the registry took a position that exposed essential cash to such a claim. The court has not decided the merits, but the budget has already changed behaviour.
The freeze also demonstrates why legal budgets need shock planning. A registry that manages scarce resources should assume high-value claims are possible. It should not be surprised that a holder facing resource withdrawal seeks large damages or security. It should ask in advance how operating accounts, strategic reserves, insurance, vendor commitments, payroll, essential service funds and court-response authority would survive an attachment or restraint order. If all money sits in ordinary accounts exposed to the same litigation pressure, the registry's continuity is legally fragile even if the balance sheet looks healthy.
That is not a plea to immunise registries from lawful claims. A registry should not be able to injure members and then hide behind public-function rhetoric. Courts must remain available. The design challenge is to distinguish funds needed for essential ledger continuity from funds available for ordinary corporate liability or discretionary litigation. If such a distinction is made only after an emergency, it will look self-serving. If it is made in advance through audited, member-visible rules, courts and members can treat it as part of the registry's public utility design.
The bank-freeze episode also changes incentives for future litigants. Once a litigant sees that freezing or threatening operational cash can move the institution, others learn from the tactic. Even if a later court rejects the approach, the signal remains. Conversely, if the registry responds by building a larger, less transparent legal reserve, members learn that litigation shocks will be priced into future fees. A provisional order can therefore produce a long-term fiscal and legal posture that outlives the immediate dispute.
Small members are least able to price this risk. A large holder can model the probability of court stress, hire counsel, negotiate indemnities and diversify revenue. A small ISP, university network or local data centre mostly needs the registry to keep records stable. It cannot influence litigation strategy, yet its operational environment is affected if the registry's accounts are frozen, staff attention is diverted, or future fees rise to rebuild legal capacity. The legal budget of a few actors becomes a quasi-insurance premium for the many.
This is why the legal-budget question is different from the reserve question. The reserve question asks how much cash is needed and how it should be protected. The legal-budget question asks what incentives the cash creates before it is spent. A large reserve may deter frivolous claims because the registry can defend itself. It may also invite larger claims because litigants see recoverable funds. It may discipline management by giving time for careful settlement. It may also let management resist settlement too long. The size of the cash pile is not the answer. The permitted use of legal money is.
AFRINIC's bank-freeze shock should therefore be treated as a pricing event. It priced the cost of severe enforcement, the leverage of provisional relief, the vulnerability of ordinary members to litigation between others, and the need for a legal budget strong enough to protect continuity but not so opaque that it becomes an all-purpose war fund.
Receivership and the court as budget governor
Receivership changes legal-budget incentives because it changes who can approve spending, who can define continuity and who can claim legitimacy for emergency decisions. In September 2023, the Number Resource Organization's public statement described the Mauritian court's appointment of an Official Receiver for AFRINIC and said the receiver's role included maintaining the status quo of assets, preserving the value of the business, overseeing elections, facilitating formation of a proper board and appointing a chief executive. The statement welcomed the development as a path toward functional governance. That description is useful as a factual exhibit. It is not the end of the analysis.
A receiver is a remedy for governance failure, not a normal operating model for a registry. During receivership, legal spending can become more legitimate in one sense and more opaque in another. It is more legitimate when it is court-supervised and directed at preserving the institution, arranging elections and keeping essential services alive. It is more opaque because ordinary member control is weakened. Members may not know how much is being spent on court applications, election vendors, counsel, communications, bylaw advice, staff matters or responses to litigants. The court sees the case file. The member sees a registry that still requires payment and trust.
The receiver's mandate also creates incentives for litigants. If a litigant believes the receiver is making decisions that affect board composition, member status, election rules or legal posture, it may challenge the receiver rather than only the registry. If a registry faction believes the receiver provides protection from a hostile member, it may seek to push decisions through the receiver's authority. If outside bodies believe the receiver is not transparent enough, they may intervene or threaten compliance processes. Legal budgets then surround the emergency bridge itself.
AFRINIC's election process showed this. The Register reported in April 2025 that AFRINIC had been unable to appoint a chief executive or elect board members since 2022, that more than twenty lawsuits were involved in the broader mess, and that a receiver was arranging elections. The same report noted concerns about potential interference, credential solicitation and the need for senior lawyers to oversee nominations. In June 2025, legal and procedural challenges delayed voting, and ICANN's attempt to change election oversight failed in court. Later that month the election was suspended and annulled after concerns over powers of attorney and voter documentation.
Each legal step had a cost and an incentive effect. Court involvement could protect fairness. It could also delay restoration. A receiver could preserve continuity. The receiver's decisions could also become new targets. ICANN could seek transparency. Its intervention could also become part of the election's contested narrative. Members could demand a clean vote. A failed vote could increase appetite for more legal controls. The legal budget was not a side expense. It was the medium through which legitimacy was being rebuilt or contested.
The key discipline during receivership should be narrowness. Legal spending should be tied to preserving essential registry services, verifying member authority, conducting elections, complying with court orders, maintaining staff capacity and preventing irreversible harm to the ledger. It should not be used to settle broad policy fights by emergency process, entrench a faction, or turn a temporary court bridge into a substitute constitution. The more abnormal the governance setting, the more constrained the legal mandate should be.
Receivership also shows why official statements about continuity cannot be treated as sufficient reassurance. NRO, ICANN, courts and registry officials may all say continuity is the goal. That may be true. But continuity has several meanings. It can mean keeping RPKI, reverse DNS, WHOIS, RDAP and allocation records functioning. It can mean preserving the corporate body. It can mean preserving current policy commitments. It can mean preventing a litigant from gaining leverage. It can mean preventing outside bodies from imposing a replacement. These meanings overlap, but they are not identical. Legal budgets should identify which continuity they fund.
The court can preserve a registry from collapse. It cannot supply the same accountability as a trusted member-governed institution. AFRINIC's receivership was therefore an emergency test of legal-budget discipline. The receiver could spend to keep the ledger alive and return governance to members. The danger was that every contested decision in that return path could require more counsel, more applications, more delay and more costs that members would ultimately bear. In that cycle, law becomes not only a remedy for governance failure but also a reason governance remains expensive to restore.
Election legitimacy has legal-budget consequences
Election disputes are not only questions of representation. In a scarcity-era registry, they are legal-budget events. A board controls budgets, appoints executives, oversees litigation, approves policy implementation, communicates risk, and decides whether to appeal, settle or reframe disputes. If the board's authority is challenged, every major legal spending decision becomes easier to attack. If an election is annulled, the cost of the failed process is not limited to ballot logistics. It includes lawyers, court time, member confusion, delayed decisions, external interventions and future risk premiums.
AFRINIC's 2025 election stress makes this concrete. Voting was scheduled after years without a functioning board. The Register reported that the June election was suspended minutes before the voting period ended because of questions regarding the validity of powers of attorney or powers given by members to delegates. South Africa's Internet Service Providers' Association alleged that authorised representatives found votes had already been submitted on their behalf through powers of attorney they had not provided, and that election officials could not produce a relevant document in another case. AFStar also alleged fraudulent powers of attorney. The receiver later annulled the election, citing concerns related to voter documentation and the need to preserve transparency and fairness.
Those were allegations and institutional responses, not final determinations of every fact. The legal-budget lesson does not require deciding each allegation. It is enough to see the cost of uncertainty. If member authority is unclear for voting, it may be unclear for litigation instructions, bylaw approvals, transfer requests, account changes, resource documents or settlement ratification. Counsel must then advise on verification. Courts may be asked to intervene. Election rules may be rewritten or defended. Members may demand investigation. The registry spends money proving who may speak for the people who fund it.
That spending can be necessary. A dirty election is more expensive in the long run than a careful one. But the legal budget also changes incentives around elections. Actors who want to delay a board can challenge procedure. Actors who want a board seated quickly can minimise irregularities. Actors who want external intervention can emphasise systemic failure. Actors who want to avoid external intervention can frame problems as isolated. Each position has legal costs and legal benefits. The registry's money, and sometimes members' private money, funds the contest over which interpretation prevails.
The September 2025 election that produced eight directors changed the immediate condition but not the incentive problem. The Register reported that seven of the eight elected directors had the endorsement of Smart Africa, that the new board could convene, and that it faced active critics, court risk, government investigation and discomfort among some community members about the concentration of support. A board can be legally constituted and still carry a legitimacy discount if members believe the path to office was contested or if litigation may constrain decisions. That discount affects the board's legal budget. It may spend more to prove authority. It may avoid settlement for fear of appearing weak. It may spend on communications to signal normalcy. It may approve reforms quickly to show momentum. Each move has incentive content.
For a registry recovering from crisis, the first legal-budget duty of a board is restraint. It should not treat election victory as a blank cheque for litigation. It should publish enough legal-spend classification to show that it is funding continuity rather than vendetta, member verification rather than factional control, and bylaw repair rather than procedural entrenchment. It should also define which legal actions require extraordinary approval because they could affect the entire fee base or the market value of member resources.
Election legitimacy also affects settlement finality. A settlement reached by a contested board may not settle the market if opponents argue that the board lacked authority, exceeded mandate or traded away member interests. A court judgment may decide a legal claim but leave the community divided if members believe the institution bought the judgment with money they could not scrutinise. Legal budgets therefore need democratic legitimacy not as a moral ornament but as transaction infrastructure. Without it, the result of litigation is discounted.
AFRINIC's election history shows that the legal budget begins long before a lawsuit caption is filed. It begins in the member register, credential controls, proxy rules, powers of attorney, nomination procedures, court-supervised election orders and the willingness to spend on verification before voting rather than on litigation after collapse. A registry that skimps on authority verification may pay far more later in legal fees. A registry that over-legalises elections may make participation so costly that only organised factions remain. The discipline lies between those errors: enough legal structure to make authority real, not so much legal machinery that the registry becomes governable only by those who can afford counsel.
The member-funded litigation dilemma
Member funding creates a dilemma no accounting label can solve. AFRINIC needs revenue to defend itself. The members who provide that revenue may disagree about the defence. Some may want aggressive action against address-market actors. Some may fear that aggressive action threatens their own reliance interests. Some may prioritise regional policy. Some may prioritise transfer liquidity. Some may want outside bodies to intervene. Some may see outside intervention as capture by another layer of authority. The legal budget aggregates these conflicting preferences into one institutional spend.
This is not unique to AFRINIC, but the African registry's crisis makes it unusually visible. In March 2026, The Register reported that AFRINIC accused Cloud Innovation, Larus and associated advocacy campaigns of creating a "web of litigation and procedural roadblocks" that delayed restoration, increased legal costs and obstructed initiatives such as training and research. The same report quoted Cloud Innovation's chief executive arguing that the dispute reflected a structural problem: the current registry model concentrates high-consequence power over economically critical internet number resources while disconnecting that power from commensurate legal and financial liability. These are litigation positions, but they identify the same funding channel. Legal conflict consumes institutional capacity that members finance and depend on.
The registry's complaint is intuitive. If one member's litigation prevents allocations, delays bylaw repair, consumes staff time and forces repeated court responses, other members suffer. They may have no stake in the disputed business model. They may simply need tickets processed, reverse DNS maintained, RPKI stable and records accurate. From this view, legal spending is a defence of the whole membership against a private actor imposing public costs.
The member-side critique is also intuitive. If a registry can threaten economically destructive action while liability is limited and legal defence is funded by the membership, the registry may not internalise the harm its discretion can cause. A holder whose business depends on address recognition may see court action as the only meaningful check on a monopoly ledger. From this view, legal spending by the member is accountability, while legal spending by the registry is institutional self-protection.
Both stories can be partly true. That is why legal-budget governance cannot be based on choosing a hero. It must be based on externalities. Which legal step protects unrelated members from service risk? Which step imposes cost on unrelated members? Which step clarifies a rule for everyone? Which step merely increases bargaining pressure? Which step repairs the ledger? Which step expands discretion? Which step preserves settlement finality? Which step makes finality harder?
The dilemma is especially sharp for small members. They rarely create the legal conflict, yet they help fund the institutional response and bear the risk of reduced service. They may lack the time to follow Mauritian proceedings, election communiques, corporate-law arguments, RIR coordination policy, bylaw consultations and transfer-policy debates. Legal spending is therefore not only a financial burden. It is an information burden. The more complex the legal environment becomes, the more influence shifts to actors with counsel, procedural memory and time.
The registry should therefore publish legal-budget categories in plain institutional language. Members do not need every pleading. They need to know, for example, how much legal spending concerns essential service defence, how much concerns the Cloud Innovation family of disputes, how much concerns election process, how much concerns winding-up response, how much concerns bylaw repair, how much concerns transfer or leasing communications, and how much concerns ordinary corporate matters. They need cost bands and reasons, not confidential tactics.
Such disclosure would not eliminate disagreement. It would make disagreement more disciplined. A member who opposes registry litigation could point to a category and argue it is discretionary. A member who supports litigation could point to a category and argue it protects continuity. The board could be judged on budget allocation rather than slogans. Courts and outside bodies could see whether the registry is preserving essential functions or financing escalation. The legal budget would remain contested, but it would stop being a fog.
The alternative is a compulsory legal levy without a constitution. In a monopoly ledger, that is corrosive. Members may continue to pay because they must, but payment without trust is not legitimacy. It is lock-in.
Settlement finality and the cost of never ending
Legal budgets change incentives not only at the start of litigation but also at the point when settlement becomes possible. An institution with enough money to continue may reject settlement terms that would have protected continuity at lower cost. A litigant with enough money to continue may reject narrow relief because broader pressure creates leverage. Each side fears that compromise will be read as weakness or precedent. The result is a dispute that remains alive after individual orders, elections or public statements suggest progress.
AFRINIC has lived inside this problem. A board was restored, but litigation did not vanish. A receiver preserved continuity, but the receiver's processes became contested. A court could allow an election to proceed, but a later election could be annulled. A court could permit ICANN to intervene in a winding-up matter, but that did not resolve the underlying dispute over Cloud Innovation, registry authority or the treatment of number resources. Public statements could announce recovery, while later statements could describe continuing litigation and legal costs. The institution moved, but finality remained expensive.
Settlement finality is harder in a registry than in an ordinary contract dispute because the settlement must satisfy more than the parties. If AFRINIC settles too narrowly with one holder, other members may ask whether a private actor captured favourable treatment. If it settles too broadly, it may appear to change policy without the community process. If it refuses settlement, members pay for continued litigation. If it settles without transparency, counterparties discount the result. If it litigates to judgment, the losing side may challenge enforcement, authority or later policy consequences. The legal budget must therefore be judged by whether it buys finality, not merely whether it buys activity.
Finality has several elements. The ledger must remain accurate. The status of disputed resources must be clear enough for customers and counterparties. Third-party network continuity should not be sacrificed to punish the direct party. Transfer and leasing implications should be stated without overclaim. Court orders should be described precisely. Confidential settlement terms should not create hidden policy. The board or authorised body should have clear power to settle. Members should be told the financial consequences in aggregate. If reserves or future fees will absorb the cost, the replenishment path should be disclosed.
These elements are not idealism. They are market infrastructure. IPv4 transactions, customer contracts and network plans need to know when a dispute is over. A settlement that ends a lawsuit but leaves resource recognition ambiguous is not final in economic terms. A judgment that affirms the registry but leaves members fearing discretionary retaliation is not final in trust terms. A court order that protects a holder but leaves the registry financially crippled is not final in continuity terms. Legal budgets should therefore be measured by reduction in uncertainty, not by number of filings.
This measurement would change incentives. Before authorising a major legal step, a board would ask: does this step make a final, member-wide settlement more likely or less likely? Does it clarify the boundary between ledger maintenance and economic control? Does it preserve unrelated member services? Does it reduce the chance that future transfer and leasing transactions require legal interpretation? Does it protect court credibility by avoiding overstatement? These questions do not predetermine the answer. They force the legal budget to serve the registry's utility function.
They also discipline member litigants. A holder seeking relief should be asked whether the relief it seeks improves finality or merely increases bargaining pressure. An injunction preventing irreversible withdrawal may protect finality by preserving the status quo. A broad damages claim, bank restraint or winding-up application may be lawful in context, but it should be judged by whether it protects the ledger or uses the ledger's fragility as leverage. The same standard should apply to both sides: litigation is justified when it narrows uncertainty around rights, records and continuity; it is suspect when it multiplies uncertainty to improve negotiating position.
AFRINIC's legal-budget incentives have often pushed in the wrong direction because the stakes are existential for the parties and diffuse for everyone else. The registry fears collapse, capture or loss of authority. Cloud Innovation fears loss of resources and business continuity. Outside bodies fear a precedent for RIR failure. Members fear service risk but have limited control. Each actor has reason to continue. Finality requires someone to price the cost of continuation for those not at the table.
That someone should be the board when the board is trusted, the receiver when the board is absent, and the court when continuity is at risk. But each must work from a narrow legal-budget constitution. Without it, settlement remains hostage to whoever can afford the next motion.
Transfers, leasing and the legal price of recognition
Transfer and leasing markets reveal legal-budget incentives because they depend on recognition rather than physical possession. An IPv4 block is useful when networks, customers, counterparties and registries accept that the relevant party can rely on it. The registry record is not the whole of that acceptance, but it is a central part. When AFRINIC litigates or communicates about transfers and leasing, it affects the market for recognition. Legal budgets become part of the cost of doing business in scarce addresses.
AFRINIC's official fee materials say that transfers between existing resource members do not, at least in the referenced schedule, attract IP or ASN transfer fees, but all organisations involved must be in good standing and transfers must comply with the applicable policy or guideline. The policy manual contains regional transfer provisions and recordkeeping requirements. These are administrative facts. Under scarcity, they become market terms. Good standing, documentation, recipient status, policy interpretation and registry timing all affect liquidity. Legal uncertainty makes each term more expensive.
Leasing adds another layer. A holder may lease addresses to customers without transferring registry title or resource registration. The customer cares about route-origin authority, reverse DNS, reputation, abuse handling, service continuity and whether the registry might challenge the upstream holder's right to provide the service. The registry cares about whether the arrangement is consistent with policy, justified need, regional purpose, accurate contact information and abuse accountability. The legal budget funds the contest over which concerns become enforceable conditions and which remain private commercial matters.
AFRINIC's public posture has often treated regional purpose and policy compliance as important constraints. Cloud Innovation and associated voices have often framed registry control over leasing and out-of-region use as overreach into commercial activity. Neither frame can be accepted as a conclusion without analysis. The economic question is what legal spending does to the boundary. If the registry can fund long litigation against leasing models, it may chill the market before a final decision. If leasing companies can fund public and legal challenges to registry communications, they may market ambiguity as permission. The boundary then emerges from legal stamina rather than from clear policy.
This is bad market design. Scarce-resource markets need boring settlement rules. A buyer should know what documents are needed, what good standing means, how long recognition takes, what disputes suspend processing, whether out-of-region use matters, how RPKI and reverse DNS are handled, and what happens if a court order intervenes. A lessee should know who is responsible for abuse contact, route-origin records, reverse DNS, customer continuity and claims of authorisation. A registry should know when it is verifying accuracy and when it is regulating a business model. Counsel should not be the only people who can answer.
Legal budgets make bad design tolerable for large actors and intolerable for small ones. A large address business can pay lawyers to navigate ambiguity. A small operator cannot. If the registry's rules are clear, legal cost falls. If the rules are ambiguous, legal cost becomes a barrier to entry and a source of bargaining power. The institution may say it is protecting the region. The economic effect may be to reward those with the money to litigate and punish those who need a predictable ledger.
AFRINIC's transfer and leasing controversies therefore belong inside the legal-budget discussion, not only inside policy debates. A transfer rule that is unclear will generate legal spend. A leasing prohibition or permission that is implied rather than written will generate legal spend. A court order that is publicly interpreted beyond its terms will generate legal spend. A member-status dispute under company law will generate legal spend. The budget is the symptom of unclear recognition architecture.
The reform is not to let every address become a freely tradable commodity without responsibilities. Nor is it to let the registry treat every commercial use as suspect. The reform is to define recognition categories tightly enough that legal budgets are not required for ordinary transactions. Fraud, false authority and inaccurate records should trigger strong action. Routine commercial delegation should trigger transparent responsibility rules. Transfers should be processed under published criteria. Out-of-region use should be addressed through clear policy if it is to be restricted, not through retrospective threat letters whose defence then consumes member money.
AFRINIC's crisis shows that legal budgets rise when recognition is vague. In a post-exhaustion registry, that is an avoidable tax.
Corruption repair without legal addiction
The reported address-record corruption history is central to AFRINIC's legal incentives because it created a legitimate demand for repair. A registry that has faced allegations of insider-linked manipulation of dormant or defunct resources cannot simply promise better culture. It must harden controls: authority verification, change logs, dual approval for high-risk records, dormant-resource review, conflict disclosures, whistleblowing channels, independent audits, staff-access controls and member notifications. Legal advice is needed for many of these steps. The problem begins when repair becomes addiction to legal control.
After a public integrity scandal, institutional psychology shifts. The registry wants to prove it is no longer weak. Staff worry about being blamed for inaction. Board members want visible reform. External critics press for cleanup. Counsel may advise that failure to review records creates future exposure. These pressures can produce necessary systems. They can also produce a habit of treating broad discretion as the antidote to past weakness. The more lawyers are involved, the more every ambiguity can be transformed into a legal risk requiring institutional control.
AFRINIC's history warns against that leap. If dormant resources were allegedly misappropriated through false authority or insider relationships, the remedy is stronger proof of authority and auditability. It does not automatically follow that every holder's evolving business model should be subject to open-ended re-justification. If old records were vulnerable, the remedy is chain-of-control verification. It does not automatically follow that the registry should gain broad power to decide whether customers are in the right geography. If staff controls were weak, the remedy is internal governance. It does not automatically follow that severe resource withdrawal is the first answer to commercial disagreement.
Legal budgets matter because they can keep the categories separate or merge them. A disciplined legal budget funds forensic review, control design, evidence preservation, court orders against actual fraud, and member-visible remediation. An undisciplined legal budget funds a generalised posture of suspicion. It treats counsel as the permanent interface between the registry and holders. It makes every resource file a potential case. It creates incentives for members to lawyer up even for ordinary interactions.
The economic cost is not only fees. It is trust. Members need to believe that the registry can correct corruption without turning correction into discretionary policing. Buyers need to believe that historical records can be verified without being reopened arbitrarily. Small operators need to believe that documentation requests will not become existential threats. Staff need to believe that following process protects them. Courts need to see that the registry distinguishes fraud from policy disagreement. If these distinctions are absent, legal budgets will expand because no one trusts the administrative path.
Corruption repair also creates communication risk. A registry may be tempted to describe enforcement as cleaning up abuse because that language attracts support. A target may be tempted to describe every enforcement step as institutional revenge because that language attracts sympathy. Both may be partly wrong. Legal-budget discipline should require the registry to state which category a matter occupies: suspected false authority, payment default, record inaccuracy, policy interpretation, customer-use dispute, transfer documentation, court compliance or service continuity. The category should determine the remedy and the legal spend.
The board's role, when functional, is to keep counsel from becoming the institution's substitute conscience. Lawyers can advise on risk, procedure and argument. Directors must ask whether the proposed legal strategy serves the narrow ledger or merely protects the institution's reputation. During receivership, that question should be asked by the receiver under court supervision and later reported to members. During staff-led operations, escalation rules should prevent individual managers from converting integrity repair into unilateral resource threats.
AFRINIC's reported corruption history justifies serious legal capacity. It does not justify legal dependency as the ordinary mode of governance. The strongest registry is not the one that can afford the most litigation. It is the one whose controls are clear enough that litigation becomes exceptional.
Outside intervention and expected rescue
AFRINIC's crisis has repeatedly attracted outside attention from ICANN, the NRO, other registries, operator groups and regional organisations. That attention can protect continuity. It can also change legal-budget incentives. If outside bodies are expected to support the registry financially, politically or legally, the registry may feel less immediate pressure to settle. If litigants believe outside bodies will intervene to preserve the registry regardless of its mistakes, they may escalate to force a broader reckoning. If courts see the registry as carrying a regional or global function, they may treat ordinary corporate remedies differently. Each effect matters before the merits are decided.
The NRO's receiver statement framed the appointment of a receiver as a path toward restored governance and continued registry services. ICANN later raised concerns about election integrity, warned of possible compliance review, and in 2026 successfully applied to become a party to Cloud Innovation's winding-up application. The Register reported that ICANN said its purpose was to help the court understand AFRINIC's unique role and to make clear that numbering resources allocated through AFRINIC are not assets of AFRINIC available for distribution in a winding-up. These are important factual interventions. They also affect bargaining.
For AFRINIC, outside support can create confidence that the institution will not be allowed to fail easily. That confidence may be necessary: a regional registry should not collapse because one dispute overwhelms it. But expected rescue can also dull discipline if management or a board believes continuity concerns will bring help even after poor decisions. For Cloud Innovation and other critics, outside intervention can look like the registry system protecting itself. That perception may encourage broader litigation or public campaigns because ordinary member processes appear insufficient. For smaller members, outside support is reassuring and disempowering at once: it may keep services alive, but it may also move the decisive conversation away from them.
Legal budgets sit inside this triangle. If the registry expects external bodies to file letters, intervene, provide support or help shape policy responses, it may spend differently. If a litigant expects external bodies to oppose remedies such as winding up, it may frame claims to challenge the legitimacy of the whole registry model rather than merely the immediate dispute. If courts expect external expertise, legal proceedings become a forum for the institutional design of the RIR system. The case becomes larger than the parties, and the legal budget expands accordingly.
Outside intervention should therefore be tied to continuity, not institutional vindication. It is one thing for ICANN or the NRO to explain that numbering resources are not ordinary corporate assets and that registry services must continue. It is another for outside bodies to be perceived as blessing every registry decision. The first protects the ledger. The second risks insulating the institution. The line matters because legal budgets respond to expected allies.
A healthy registry should welcome factual assistance and reject blank-cheque support. It should be able to say: outside bodies may help protect essential services, explain technical consequences, provide emergency coordination and support lawful governance restoration; they should not provide cover for discretionary legal escalation or policy shortcuts. That position would reduce the moral hazard created by expected rescue.
AFRINIC is a hard case because a failing registry really can create regional and global coordination risk. But the answer to that risk cannot be that the registry's legal budget becomes morally untouchable. The more essential the function, the stronger the need to classify legal spending. Public importance is not an exemption from economic discipline. It is the reason discipline must be visible.
What legal-budget discipline would require
AFRINIC's legal-budget problem cannot be solved by demanding lower spending in the abstract. Under stress, low spending may be reckless. The registry needs counsel for court orders, member disputes, fraud repair, election integrity, bylaw conflicts, winding-up risk, staff protection and continuity planning. The solution is not a cheap legal budget. It is a constitutional legal budget: one that defines purpose, authority, limits, reporting and settlement logic before the next crisis forces improvisation.
The first requirement is categorisation. Every material legal matter should be placed into a member-visible category without disclosing privileged advice: essential service continuity, resource-status dispute, record-integrity repair, election and member-authority verification, corporate governance, bylaw reform, winding-up or insolvency response, transfer and leasing communications, ordinary commercial matters, employment, debt collection and public communications. Categories matter because each has a different claim on compulsory member funding.
The second requirement is a decision memo before severe action. If the registry contemplates terminating membership, reclaiming resources, freezing a transfer, challenging a leasing model, appealing a major order or resisting settlement beyond a defined cost band, the decision should identify the ledger harm, the evidence, the policy or contractual basis, the proposed remedy, narrower alternatives, expected legal cost, service-continuity risk, member-wide externalities and settlement path. This memo need not be public in full. Its existence and category should be reported. The point is to force the incentive question before counsel defends the action.
The third requirement is cost bands and triggers. Legal spending has a way of becoming inevitable after the first invoice. A registry should define thresholds at which a matter requires renewed approval, independent review or member disclosure. A case costing a small share of annual revenue is one thing. A case that could consume a large share of reserves, affect future fees or threaten service continuity is another. The board, receiver or emergency authority should not be able to let a dispute drift across those thresholds without explaining why continued spending still protects the narrow ledger.
The fourth requirement is settlement principles. The registry should state in advance that settlements must preserve uniqueness, accuracy, third-party continuity, non-discrimination, court compliance and member-visible policy boundaries. It should also state what it will not buy with settlement: hidden policy changes, private capture, retrospective punishment unrelated to ledger integrity, or confidentiality that prevents members from understanding material financial consequences. These principles would make settlement less vulnerable to accusations of weakness or favouritism.
The fifth requirement is an essential-service firewall. RPKI, reverse DNS, WHOIS, RDAP, IRR, allocation records, member support for critical changes, security monitoring and record-preservation functions should be financially and operationally insulated from litigation as far as law allows. Legal budgets should not be allowed to consume funds needed for these functions without explicit emergency approval and disclosure. A lawsuit over one member's resources should not make unrelated members wonder whether publication services will continue.
The sixth requirement is independent review before existential remedies. A registry does not surrender authority by creating a credible review path. It reduces litigation demand. If members believe the only meaningful check on resource withdrawal or severe account action is court, they will go to court. An independent review mechanism that can examine evidence, policy interpretation, proportionality and interim continuity would be cheaper than emergency proceedings and better aligned with the registry's utility function.
The seventh requirement is post-crisis reporting. After a major case, election challenge, bank restraint, receivership step or winding-up response, the registry should report aggregate legal cost, category, authority, service impact, settlement or judgment effect, lessons and control changes. The report should distinguish unavoidable defence from avoidable escalation. It should say whether future fees or reserves will be affected. Members cannot judge incentives if every legal episode disappears into a line called professional fees.
None of this requires AFRINIC to adopt the worldview of its critics. It requires it to show that its legal spending is tied to the function members cannot replace: a reliable, narrow registry ledger. The same framework would also discipline critics. A member seeking relief could be asked whether its legal action protects record certainty or exploits institutional fragility. Legal-budget discipline should bind the institution, but the standard it creates can be used to evaluate everyone who litigates around the ledger.
Continuity versus self-protection
The hardest distinction in AFRINIC's crisis is between protecting registry continuity and financing institutional self-protection. The two often look similar in the moment. A lawsuit threatens bank accounts; the registry hires counsel. Is that continuity or self-protection? A member seeks winding up; the registry resists. Is that continuity or self-protection? A board faces claims about election legitimacy; it pays lawyers to defend its authority. Is that continuity or self-protection? A public statement challenges a leasing claim; counsel reviews the wording. The answer depends on the function being protected.
Continuity means the ledger remains accurate, available and neutral while lawful disputes proceed. It means RPKI material is not used as leverage. Reverse DNS does not become collateral damage. WHOIS and RDAP records remain reliable. Allocation and transfer records are preserved. Member authority is verified. Court orders are implemented precisely. Staff can perform essential duties. Members receive factual risk information. The institution has enough money and authority to survive while governance is repaired or while a court decides.
Self-protection begins when the object shifts from the ledger to the institution's freedom from discipline. It appears when legal spending is used to defend broad discretion without explaining the ledger harm. It appears when communications aim more at delegitimising critics than informing members. It appears when every demand for disclosure is treated as an attack. It appears when settlement is rejected because compromise would reduce institutional pride rather than because it would harm the registry. It appears when continuity language is used to fund legal positions that members would contest if the costs were separately shown.
AFRINIC has arguments for continuity. It serves 54 countries across Africa and the Indian Ocean. Its services are hard to replace quickly. The region's operators need stable records. The reported address-record scandal showed that the ledger can be harmed by weak controls. The Cloud Innovation dispute and bank freeze showed that one conflict can threaten the institution. Receivership and election repair were directed at restoring governance. Winding-up risk required a response because number resources should not be treated as ordinary corporate assets. These are serious continuity claims.
AFRINIC's critics also have arguments against self-protection. A registry with monopoly control over recognition can harm members before a court supplies a remedy. Legal liability may be limited relative to economic impact. Aggressive enforcement can threaten customers and business continuity. Board and election legitimacy problems can make legal spending suspect. Transfer and leasing restrictions can reduce liquidity and increase dependence on the registry. Public-function language can be used to resist accountability. These are serious discipline claims.
The value of legal-budget analysis is that it does not need to dismiss either side. It asks what each dollar is buying. If the dollar keeps RDAP, WHOIS, RPKI, reverse DNS and allocation records stable during court stress, it is continuity spending. If it verifies member authority before an election, it is continuity spending. If it preserves records during a fraud investigation, it is continuity spending. If it resists a winding-up theory that would treat number resources as distributable assets, it is likely continuity spending, though still subject to cost discipline.
If the dollar funds an attempt to make disputed policy broader than its text, it is suspect. If it sustains a refusal to disclose aggregate legal categories, it is suspect. If it converts a communications dispute into a reputational campaign, it is suspect. If it postpones settlement without a member-visible theory of finality, it is suspect. If it lets officials avoid explaining why a severe remedy was chosen over narrower alternatives, it is suspect.
This distinction should be built into AFRINIC's recovery. A restored board and budget are not enough. The institution must show that legal spending will no longer be a general shield around whoever controls the registry. It must show that the legal budget is subordinated to the ledger, not the other way round. If it can do that, counsel becomes a tool of continuity. If it cannot, counsel becomes the institution's most expensive substitute for legitimacy.
The lesson of AFRINIC's invoice
AFRINIC is not merely a cautionary story about one region or one dispute. It is a test of how monopoly registry institutions behave when the resources they administer become scarce, valuable and legally contested. The formal language of internet number governance was built around coordination, uniqueness, need, stewardship and community process. The modern economic environment adds asset value, leasing revenue, transfer markets, bankable reliance, litigation finance, court supervision and member dependence. Legal budgets are where the old language meets the new incentives.
The invoice tells members whether the institution understands that change. A narrow registry spends on lawyers to keep the ledger accurate, services stable, fraud contained, court orders complied with, elections credible and settlements final. A self-protective registry spends on lawyers to defend its discretion, delay accountability, control narratives and make members finance the cost of institutional pride. In practice, the same institution may do both in the same year. That is why classification matters more than rhetoric.
AFRINIC's recent history contains every warning sign: reported record corruption, scarcity-era enforcement, high-value litigation, provisional account freezes, board absence, receivership, election legitimacy stress, winding-up risk, transfer and leasing pressure, outside intervention, and member dependence on services that cannot be replicated by choosing another vendor. It also contains every reason not to oversimplify. A registry can be attacked unfairly. A litigant can be threatened unfairly. A court can protect rights and still create systemic risk. Outside bodies can preserve continuity and still appear to protect incumbency. Members can demand accountability and still underestimate the cost of keeping a critical ledger alive.
The discipline AFRINIC needs is modest in theory and hard in practice. Legal money should be available, but not unclassified. Enforcement should be possible, but not legally subsidised into maximalism. Court access should remain open, but remedies should be judged by their effect on unrelated members. Receivership should preserve services, not become a new site of policy power. Elections should restore authority, not provide a blank cheque for legal spending. Transfer and leasing rules should be clear enough that ordinary market actors do not need litigation to know what recognition means.
Above all, AFRINIC should treat legal spending as a form of institutional power. It should be authorised, limited, disclosed in category, reviewed after major matters and tied to settlement finality. Members should not be asked to trust that every invoice protects them. They should be shown, in aggregate and in plain language, which risk the invoice addressed and why the chosen legal path was cheaper for the ledger than the alternatives.
That standard would not end AFRINIC's disputes. It would change their incentives. Management would know that severe action carries a pre-commitment to explain cost and proportionality. Litigants would know that the registry can defend continuity without hiding discretionary escalation. Courts would see which functions are essential and which are ordinary corporate interests. Members would see whether their fees support the ledger or the institution's self-image. Markets would price AFRINIC records with less guesswork.
The opening bill is therefore the whole story in miniature. It can be the price of defending the shared address book. It can be the price of delaying accountability. In a monopoly registry under IPv4 scarcity, no one should have to infer the difference from the size of the cheque.