Summary

  • The persons AFRINIC described as its Board and the still-undischarged receiver announced a Legal Committee on 13 October 2025, with Ben Roberts as chair and a mandate to inventory ongoing cases, optimise legal resources and maintain a coherent strategy. The announcement proves those words were issued; it does not prove the purported Board’s complete authority chain, the committee’s valid constitution or its performance.
  • The checked public record does not disclose the creating resolution, terms of reference, full membership, delegation, conflicts, minutes, versioned matter baseline, counsel and retainer register, case budgets, invoice and payment mapping, settlement criteria or completed report. Those are not ornamental disclosures. They are the controls that distinguish a committee from an additional layer of untraceable agency.
  • AFRINIC’s court index can seed a useful inventory, but its identifiers, parties, descriptions, statuses and remarks do not connect each matter to authority, counsel, cost, exposure, operational risk, objective, deadline, outcome or exit condition. A case-to-authority-to-cost-to-outcome ledger is the minimum credible product.
  • Confidentiality is a strong argument for protecting privileged advice, evidence, personal data and live negotiating positions. It is not an argument for keeping secret the committee’s authority, composition, conflicts, matter-count method, aggregate spend, approval chain, governing criteria or reporting line.
  • AFRINIC remains a narrow registry ledger, bookkeeper and coordinator. It has no sovereignty or punishment power. Legal disputes belong to ordinary corporate authority, contract and competent courts, while WHOIS, RDAP, RPKI and registration services must remain continuous and non-retaliatory.

A chair without a visible charter

The human scale of AFRINIC’s legal problem appears in one name. On 13 October 2025 Ben Roberts was publicly named as chair of a new Legal Committee. The appointment placed a person at the centre of a docket that AFRINIC later said had cost millions of dollars and continued to complicate its recovery. Yet the public record checked through 10 August 2026 does not show the document that tells Roberts what his chair actually permits him to do.

Can the committee merely advise? Can it instruct a lawyer, open a matter, renew a retainer, challenge an invoice, recommend a settlement, authorise an appeal or discontinue a case? Who else sits on it? Who records its decisions? When does its mandate expire? Which matters remain reserved to the purported Board, which require the receiver’s participation and which belong to an authorised officer? If a member has an interest in a lawyer, a party or the subject of a dispute, where is that conflict declared and how is recusal recorded? The announcement supplies none of these answers.

That does not prove that the documents do not exist. The absence is bounded: no creating resolution, terms of reference, complete roster, minutes, conflict register or finished report was located on the checked public surfaces by the cutoff. Some records could be private, unindexed or unavailable. But the distinction between “not publicly located” and “non-existent” does not make the gap unimportant. An institution asking members to trust a committee can publish enough governance metadata to show that the committee has a lawful shape without disclosing privileged advice.

The announcement itself was joint. It came from the persons AFRINIC described as its Board and Gowtamsingh Dabee, the court-appointed receiver. Five days earlier, the receiver had applied for termination of the receivership. The communique nevertheless said formal discharge was still awaited and that he would continue supporting AFRINIC. AFRINIC’s own update of 12 March 2026 again said discharge remained pending and acknowledged proceedings challenging the directors’ appointments and other governance acts. NRS disputes the current body’s authority.

The analytically correct description is therefore the purported Board: neither the election announcement, continued operation, a company filing nor the Legal Committee communique is a final judicial validation of every act.

This matters because a committee is an agent. It cannot acquire more authority than a lawful principal has delegated, and it cannot cure a defect in that principal by acting busily or competently. The receiver’s participation may be valuable for continuity, custody and access to records. It does not silently settle whether the purported Board, receiver, committee or an officer possessed the particular power required for each instruction, retainer, invoice, payment, appeal or settlement. The authority question must be answered transaction by transaction, at the date of the act.

Three ambitions are not terms of reference

The communique gave the committee three intelligible objectives: take stock of all ongoing cases involving AFRINIC, optimise legal resources and maintain an appropriate and coherent strategy. Consolidation is sensible. A scattered docket creates duplicated instructions, missed deadlines, inconsistent positions and invoices that no one can reconcile to a single institutional purpose. A small committee can reduce the burden on a full board and develop the expertise needed to compare cases across counsel and courts.

But an objective is not a power, and a mandate announced in prose is not a full charter. AFRINIC’s 2016 bylaws contemplate committees appointed for reasons and with terms of reference the directors consider necessary or desirable. They separately address delegation, conflicts, quorum, voting, minutes and written resolutions. They also place budgets, expenditure ceilings, execution and payment authority on distinct control surfaces. The Mauritius Companies Act supplies the domestic corporate-law setting in which those instruments and acts must be assessed.

The existence of a general power to form committees therefore sharpens the documentary question; it does not dissolve it.

A creating resolution should identify the meeting or valid written process, the participating decision-makers, quorum, vote, signatures and minute entry. A charter should then specify purpose, composition, chair, secretary, term, reporting line, meeting cadence and access to records. Most importantly, it should draw the boundary between advice and decision. If the committee may exercise a delegated Board power, the instrument should name that power, its limits and expiry. If it cannot, the record should say who converts a recommendation into an effective client instruction.

Reserved matters deserve particular clarity. Retaining or terminating counsel can change both strategy and future cost. Approving a budget can commit member funds before a pleading is filed. Settling, appealing or discontinuing may surrender claims, preserve rights, accept obligations or close off later options. Investigating alleged Power of Attorney or Power to Vote documents raises custody, privacy, conflict and referral questions. None of these powers should be inferred from the elastic phrase “coherent legal strategy”. Each requires an authorised decision-maker and a contemporaneous record.

The committee’s additional announced work illustrates the risk. The communique said it was reviewing information concerning allegedly fraudulent Power of Attorney or Power to Vote documents from the aborted June 2025 election and would update members. That establishes only that the issuers made the statement. It does not establish fraud, a document count, a finding, a referral or a completed investigation.

A proper process would identify the material’s custodian, preserve evidence, separate affected decision-makers from the review, protect personal information and refer questions to an independent investigator or competent authority where required. The Legal Committee cannot create guilt. Nor can a registry use technical service, membership status or number-resource administration as punishment.

The committee also needs a clear addressee. Formally, it must answer through the lawful corporate chain: AFRINIC Ltd acting through valid organs and authorised officers, with any receiver power treated according to the relevant court authority rather than implied from presence. Institutionally, it must provide bounded reporting to Resource and Registered Members whose fees, rights and operational continuity bear the consequences. Members do not become day-to-day litigation managers merely because they pay fees or vote at a meeting.

But they are entitled to know whether lawful actors committed company funds, whether conflicts were managed and whether the portfolio produced defensible results. Approval of accounts records a member act; it cannot retrospectively manufacture the contemporaneous authority that a retainer, instruction, invoice or payment needed when made.

From court list to control ledger

AFRINIC’s public court-case index is useful. It presents case references, parties, the nature of matters, status descriptions, remarks and occasional links to orders. It includes completed and ongoing proceedings and continued to show updates through at least 11 June 2026. An alternate route provides a searchable public copy of historical rows. These surfaces are better than silence because they give the committee a place to begin reconciliation.

They are not, however, a governed portfolio. The presentation has inconsistent numbering and evolving status information. A row may describe litigation, but it does not say whether the matter is duplicated elsewhere, which date governs the stated status or who signed off on the update. It does not establish that the list is complete. Most important, it does not show the institutional links on which control depends.

The committee’s first useful product should therefore be a versioned ledger with a stable matter identifier distinct from changing labels on a web page. Each entry should contain the court reference, parties and capacities, jurisdiction, procedural stage, relief sought, potential exposure, current status and an “as of” date. It should identify the next event and deadline, because strategy without a calendar is merely intention. It should also state the operational-continuity risk: whether a decision threatens routine registry services, custody, staff capacity or member access.

Those descriptive fields are only the first half. Every matter also needs an authority chain. The ledger should name AFRINIC’s accountable decision owner, identify the instrument authorising the opening or continuation of the matter and record whether the relevant authority is current, expired, reserved or contested. It should name internal and external counsel, connect them to the governing retainer and show that conflicts were checked. The budget column should distinguish an approved ceiling from spend to date and forecast cost. Variances should identify who approved them and why.

Invoices and payments require their own links. An invoice record should show matter, timekeeper, time, task, contractual rate, tax and disbursement treatment, approver and exception. The payment record should show the authorised chain, value date, account and payee reconciliation. None of this requires publication of privileged legal analysis. It requires the committee, finance function, company secretary and counsel to be able to follow a charge from work performed to invoice, approval, payment and matter outcome.

Finally, a controlled matter needs a purpose and an exit condition. What right or asset is being protected? What exposure is being reduced? Which result is realistically available? At what cost or factual change should AFRINIC settle, appeal or discontinue? Who holds that decision? A status such as “ongoing” is not a strategy. Nor is a claimed victory necessarily evidence of value if it leaves larger exposure, triggers another appeal or consumes funds without protecting registry continuity.

The committee should measure fewer unresolved authority gaps, better traceability, deadlines met, exposure avoided, rights preserved and cases responsibly exited—not the volume of filings or legal work generated.

This is why a case list becomes governance only when it connects case, authority, instruction, counsel, retainer, budget, invoice, payment, conflict, objective, settlement criteria, outcome and reporting line. Without those links, the institution records legal motion after the fact. With them, it can decide before the next cost and risk become inevitable.

A historical retainer as a test case

The disclosed C&A Law engagement offers a concrete artifact against which to test the committee’s method. The letter was dated 20 October 2021, received by AFRINIC the next day and accepted by then-chief executive Eddy Kayihura on 25 October. It said the firm would act from 1 August 2021 in twelve listed matters. It also covered ad hoc assistance and new cases. Twelve is therefore a starting baseline, not a reliable later total for AFRINIC’s docket.

The stated professional-fee term was USD 1,000 per hour, including the fees of named senior counsel, exclusive of 15% VAT and disbursements. It contained invoice, confidentiality and termination provisions, including a 30-day invoice term. Those facts matter, but so do their limits. The letter does not prove that USD 1,000 was chargeable separately for every lawyer. It does not prove a concurrent USD 2,000 rate. Nor does execution of the letter prove every later instruction, hour, invoice, payment, amendment or outcome.

A competent Legal Committee would use this retainer to demonstrate its reconciliation discipline. It would create a record for each of the twelve opening matters, then separately document every later matter admitted under the open-ended scope. Each matter would point to the person who instructed the work and to that person’s authority on that date. Amendments, termination status, competing quotations and conflicts would be recorded rather than assumed. Time and disbursements would be matched to the relevant matter.

The invoice would be tested against contract terms and approved budget; the payment would then be matched to a valid signatory chain. The outcome column would state what the work secured, prevented or concluded.

Aggregate figures show why this exercise is material while also showing why it cannot end at the accounts. NRS calculated USD 3,289,408 in disclosed legal expenditure across 2022-2025 and USD 2,148,059 paid to C&A Law in 2022-2023, drawing on AFRINIC publications. AFRINIC’s consolidated annual report disclosed USD 1,133,630 of legal expense for 2023 and USD 27,322 for 2024. AFRINIC later said litigation had cost millions. These figures establish a high-stakes need for reconciliation; they do not establish a case allocation, accounting composition, cash date, rate reasonableness, wrongdoing or value achieved.

The committee should resist two easy but opposing mistakes. The first is to treat an audited or approved aggregate as proof that every underlying engagement and payment was authorised. Accounts recognise financial events. They do not reach backward in time to create a missing mandate or cure an invalid approval. The second is to infer misconduct merely because costs are large or records have not been published. High spend demands evidence, but the fair conclusion remains documentary: show the authority, work, invoice, payment and result for each transaction.

Until that reconciliation exists, neither blanket vindication nor blanket accusation is justified.

The case for confidentiality—taken seriously

The strongest defence of the committee deserves more than a dismissive answer. AFRINIC faced a broad and costly docket while trying to move from receivership towards ordinary governance. A newly reconstituted decision-making body could reasonably prefer a small group able to read privileged advice, compare counsel, respond quickly to deadlines and negotiate without broadcasting its reserve position to opponents. Publicly revealing evidence, legal advice, settlement ranges or the vulnerabilities of witnesses could prejudice cases. Personal data in the election-material review may require careful protection.

The receiver’s participation could also maintain custody and institutional memory during an unsettled handover.

This argument gets several important things right. Fragmented instructions are usually worse than a consolidated view. Specialisation can improve decisions. Privilege and privacy are not excuses invented by lawyers; they protect the client’s ability to obtain candid advice and protect people whose data sits in contested files. Settlement leverage can be destroyed by disclosing tactical thresholds. In fast litigation, insisting that every operational decision wait for a full-member debate would be inefficient and sometimes dangerous.

The answer is not radical transparency. It is separation between protected substance and accountable metadata. Legal advice, evidentiary detail, personal data and live negotiating positions may be withheld or narrowly redacted. The committee’s creating resolution, charter, roster, term, reporting line and general delegation are different. So are conflict declarations or, where details cannot be published, the fact that a declaration was made and a recusal occurred.

A matter identifier, procedural status, responsible owner, counsel identity, aggregate spend, budget status, next public deadline and outcome can often be reported without exposing privileged merits.

The same separation applies to decision criteria. The public need not know a confidential settlement number to know whether authority belongs to the purported Board, receiver or another officer, whether independent advice is required above a threshold, whether a conflicted member is excluded and whether an outcome was tested against continuity and cost. Secrecy about advice can preserve the client’s position. Secrecy about who the client is, who may bind it and who approved payment destroys the ability to establish that there was an accountable client at all.

Efficiency also needs a denominator. A committee that produces faster instructions but cannot show lawful authority, budget control or resolved exposure may merely accelerate risk. A committee that reduces the number of advisers but concentrates unchecked influence in one professional relationship may lower visible complexity while increasing dependency. Proper metadata enables speed because officers know in advance which decision-maker owns each threshold. It is not bureaucracy added after action; it is the map that prevents action from being repeated, challenged or disowned.

Who bears the loss, and who gets paid to act?

The committee sits inside a classic agency problem. Committee members, the receiver, management, company secretary, finance staff and external lawyers all perform roles for an institution whose members and resource holders bear much of the downside. These agents are not presumed disloyal. The problem is structural: professional activity can continue, invoices can be paid and institutional roles can be preserved even when the connection between a case and the principal’s benefit is hard to test.

Legal work is particularly prone to self-reinforcement. One proceeding can create advice, applications, responses, appeals and adjacent disputes. Each step may be defensible on its own. Yet without a portfolio view, no decision-maker sees when the sequence has ceased to protect an achievable right or has begun mainly to reproduce more process. An open-ended retainer makes matter-opening discipline still more important, because “new cases” can enter a professional relationship without the public visibility of a new contract.

Traceability changes those incentives. Naming a lawful instruction owner makes it harder for responsibility to disperse across “the Board”, “management” and “counsel”. A budget ceiling requires an explicit decision when cost grows. A conflict register distinguishes expertise from dependence. An exit condition forces discussion of settlement and discontinuance before sunk costs make retreat institutionally embarrassing. An outcome field asks what changed for AFRINIC, rather than what lawyers did.

The committee must not answer primarily to the professionals whose work it commissions, to the purported Board’s need for institutional validation or to external supporters who prefer a reassuring story of regional stability. ICANN, the NRO and peer institutions may coordinate or support continuity. Their recognition cannot provide Mauritian corporate authority, member consent, a valid retainer or a payment approval. Practical endorsement sits outside the authority chain.

Members, too, require a precise place. They are affected principals and fee payers, not a sovereign “community” whose applause can replace law. Their role is to receive intelligible accounts, insist on valid governance records, elect or remove actors where lawful processes allow, and make reserved corporate decisions on adequate information. The committee should report enough for them to judge stewardship, but it should not outsource professional litigation calls to a plebiscite. Accountability means a named agent explaining an authorised decision to the principal; it does not mean every principal personally giving the instruction.

A registry cannot turn litigation into jurisdiction

The deepest boundary is easy to lose because AFRINIC performs an operationally important task. It maintains a number-registry ledger and coordinates records on which networks rely. That central position is useful, but it is not sovereignty. AFRINIC is not a legislature, regulator, prosecutor or punishment authority. Its entries and service interfaces record and coordinate operational reality; they do not author the underlying rights, create judicial power or validate a corporate act merely because the system continues to run.

This distinction bears directly on the committee. A coherent litigation strategy may protect AFRINIC as a private company. It cannot convert a committee recommendation into jurisdiction over operators, make an allegation an adjudicated fact or use access to WHOIS, RDAP, RPKI, registration or membership as leverage against litigants and unrelated resource holders. Alleged election misconduct belongs with preserved evidence, lawful investigation and competent authorities. Corporate disputes belong under the company’s instruments, contracts and court orders.

Routine service continuity should therefore be insulated from uncertainty over the purported Board, the receiver or the committee. If an authority defect is later established, the response should examine affected transactions individually, use lawful correction where available and preserve reversible options. It should not punish networks that did not authorise the disputed act. If the committee proves its authority and value, the same separation remains wise: stable ledger services should not depend on whether a litigation strategy enjoys political approval.

This is also the proper measure of institutional recovery. More filings, favourable communiques or external expressions of support do not by themselves show control. Recovery becomes credible when routine operations are boring, legal decisions are attributable, costs are reconcilable, conflicts are managed, losing options can be exited and members can see what was protected. The Legal Committee can contribute to that outcome. On the checked public record, it remains an announced control layer rather than evidence that control has been achieved.