Summary

  • AFRINIC reported USD 157,699 for 3 Hare Court, USD 55,484 for Queens Court Chambers and USD 44,605 for Georges Penny Chambers. Those exact three lines sum to USD 257,788. The disclosure establishes the labels, amounts and arithmetic, but it does not establish that there were three contracts, three invoices, three payments or three separate legal-advice retainers.
  • The receiver’s public appointment record names four Nomination Committee appointees: Simon Nicholas Davenport KC as chair, and Nicholas George Leah, Priscellia Elouise Robinson and George Hayden Penny as assistants. Their stated mandate was collective. The three-line financial table therefore cannot, on its own, show how responsibilities or costs were divided among the four people.
  • Two finance labels strongly correspond to published professional affiliations: Nicholas Leah with 3 Hare Court and Priscellia Robinson with Queens Court Chambers. The third does not reconcile cleanly. AFRINIC’s finance label says “Georges Penny Chambers,” while its FAQ and the professional register place George Penny at 187 Chambers. That is an identity and payee-control gap, not proof of false billing.
  • Multiple barristers could have been justified by urgency, parallel workload, independence, specialist judgment or availability during a court-structured effort to restore the Board. The defensible accountability question is therefore not whether three advisers were inherently improper. It is whether AFRINIC can document the mandate, procurement route, privilege boundary, conflict controls, responsibility allocation, fee formation, invoice review and payment status attached to each reported line.

Analysis

The unusually precise part of the record

This inquiry begins with what is unusually clear. AFRINIC’s financial-year webpage and its document titled “2025 Unaudited Finance Report” publish the same three named cost lines under the Board election held on 23 June 2025: USD 157,699 for 3 Hare Court, USD 55,484 for Queens Court Chambers and USD 44,605 for Georges Penny Chambers. The three add exactly to USD 257,788.

That total is not an estimate and it does not require a theory about what happened behind the table. It is a direct arithmetic consequence of AFRINIC’s own reported figures. The largest line accounts for about 61.174 per cent of the three-line sum, the Queens Court Chambers line about 21.523 per cent, and the Georges Penny Chambers line about 17.302 per cent. The largest exceeds the smallest by USD 113,094 and is about 3.535 times its size.

Those comparisons describe the numbers; they do not explain them. A larger line may reflect more hours, a broader remit, senior support, different disbursements, a different fee arrangement, currency conversion or work performed by more than one person. None of those possibilities is proved in the checked record. Equally, the size difference does not prove duplication, overcharging, impropriety, waste or an unreasonable price. There is no public denominator—no hours, rate, cap, fixed-fee scope, deliverable schedule, expense policy or acceptance record—against which a responsible price judgment could be made.

The precision of the total can therefore be misleading if it is allowed to stand in for precision about the transaction. USD 257,788 is closed arithmetic. The responsibility chain behind it remains open.

Four appointees, one collective mandate, three reported lines

The appointment record supplies the next set of firm facts. On 21 April 2025, the receiver’s communiqué appointed Simon Nicholas Davenport KC as chair of the Nomination Committee. Nicholas George Leah, Priscellia Elouise Robinson and George Hayden Penny were appointed as assistants. AFRINIC’s subsequent FAQ associated the chair with 4 Pump Court and the assistants with 3 Hare Court, Queens Court Chambers and 187 Chambers respectively.

The committee’s published duties were substantial. As a collective body, it was to design and implement the election process and timeline, call for candidates for Board seats, prescribe candidate criteria, interview candidates where appropriate, review and finalise the candidate list, and supervise polls conducted through the election apparatus. The receiver approved guidelines that the committee had drafted, according to AFRINIC’s FAQ. This was not a ceremonial appointment. The committee sat on a consequential control surface: it helped shape access to candidacy and the conduct of the process intended to restore AFRINIC’s Board.

But the public mandate speaks in the language of the committee. It does not allocate one workstream to the chair, another to a named assistant, or a particular deliverable to any chambers. It gives no responsibility matrix, matter code, work plan, time record or acceptance criterion. It does not say who carried primary responsibility for candidate criteria, interviews, list finalisation, guideline drafting or supervision. Nor does it show whether the chair’s work was separately compensated, included in another arrangement, waived or simply represented elsewhere.

That matters because the personnel record and the financial record have different shapes: four appointees on one side and three named chamber or barrister-linked cost lines on the other. The checked table contains no separate line naming Simon Davenport or 4 Pump Court. It would be wrong to turn that absence into a conclusion that the chair worked without compensation. It would also be wrong to assume that the largest line included his work, or that the three lines mapped one-for-one to the three assistants. Each is a possible explanation; none is documented in the public record available to this inquiry.

The correct formulation is narrower and more useful. AFRINIC disclosed three cost lines connected by name or professional address to a four-person committee whose duties were expressed collectively. Without the underlying engagement structure, the public cannot tell whether this was one coordinated team arrangement, several separate arrangements, a mixture of individual and chambers billing, or some other structure. “Three cost lines” is the fact. “Three contracts,” “three invoices,” “three payments” and “three retainers” are unproved reconstructions.

Two strong correspondences and one material discrepancy

The public records permit careful identity mapping, but only up to a point. The official FAQ associates Nicholas Leah with 3 Hare Court, and the current professional register does the same. That gives a strong affiliation correspondence with the USD 157,699 line. The FAQ and register similarly associate Priscellia Robinson with Queens Court Chambers, corresponding strongly to the USD 55,484 line.

An affiliation correspondence is not transaction proof. It does not identify the contracting party, establish that the chambers itself was the legal supplier, show whose tax identity appeared on an invoice, or prove that a named individual received funds. Chambers labels can operate differently from incorporated supplier names, and the public table does not settle that legal character. The safe conclusion is that the first two labels line up with the recorded professional affiliations of two assistants. The contract, payee, invoice and payment links remain unproved.

The third line is more difficult. The finance disclosure says “Georges Penny Chambers” and assigns USD 44,605 to it. The appointment record names George Hayden Penny, without the final “s” attached to his first name, while AFRINIC’s FAQ and the professional register associate him with 187 Chambers. No checked record identifies an entity called Georges Penny Chambers as the contracting party or payee. No invoice or correction notice reconciles the wording.

The probable person correspondence is evident: the committee included George Penny and the finance label resembles his name. Yet probability cannot be used to manufacture a legal entity. Silently rewriting the label as 187 Chambers would erase the very control issue the disclosure creates. Treating “Georges Penny Chambers” as a proven payee would be equally unsafe. The discrepancy might be a transcription error, shorthand, a classification convention or something else entirely. A simple reconciliation record could potentially resolve it.

Until then, the correct position is that the USD 44,605 line has a material naming gap between AFRINIC’s finance table and AFRINIC’s own appointment material.

This distinction protects all sides. It prevents a clerical mismatch from being inflated into an accusation of false billing, while also preventing a plausible guess from being presented as audited identity. Good accountability does not require the most suspicious interpretation. It requires that names attached to costs be reconcilable to the parties authorised to perform, invoice and receive payment for the work.

What the official record proves—and what it cannot carry

Official sources deserve full weight for official acts. AFRINIC’s documents prove that its receiver appointed the four people, that AFRINIC described their collective mandate and selection rationale, that an election process was conducted, that the receiver later annulled it for stated concerns involving voter documents, and that AFRINIC’s finance function later published the three labels and amounts. The professional register proves the current recorded affiliations of the three assistants. ICANN’s announcement proves that ICANN raised an institutional concern and asked for remedial action.

Those sources do not become independent verification of every implication surrounding the acts they report. The finance table does not prove contract formation or payment. The appointment communiqué does not prove the absence of conflicts. The professional register does not identify AFRINIC’s 2025 supplier or bank beneficiary. The receiver’s election mandate does not itself approve every provider, price or invoice. ICANN’s announcement establishes ICANN’s claim, not the truth of an allegation or an adjudicated conflict.

The annulment establishes the receiver’s act and stated reason, not that any professional line was valueless or caused the outcome.

Document character is especially important here. The provider detail is expressly unaudited. A later audited annual financial statement is a separate instrument operating within its stated financial-reporting scope and materiality. The existence of that audit is relevant, but it cannot be silently expanded into public proof that these exact engagements were authorised, procured under an appropriate route, conflict-cleared, accurately named, independently fee-reviewed, invoiced and paid. Unless transaction testing is expressly disclosed, annual-statement assurance and transaction-level accountability remain distinct.

Likewise, a later invitation to members to receive an auditor’s report or approve annual accounts cannot automatically create the missing ex ante transaction chain. The checked record does not close the result or legal effect of the proposed 2026 Ordinary Resolution VI. Even if members later approved accounts, the effect of that act on earlier authority, waiver or ratification would require its own record and analysis. Presentation after the event is not self-proving authority before the event.

Election authority is not a blank cheque

The election took place in exceptional circumstances. The Supreme Court of Mauritius had appointed a receiver in the AFRINIC process, and the institutional objective was Board reconstitution. The receiver then appointed the Nomination Committee and, according to AFRINIC, approved its guidelines. These acts establish a chain for creating and directing an election mechanism.

They do not, by themselves, show the financial delegation for each professional commitment. Several different decisions sit inside what may look from a distance like a single mandate: appointing a committee member, engaging a service provider, identifying the client, choosing a procurement route, setting the fee basis, approving expenses, managing conflicts, reviewing work, certifying an invoice and authorising payment. One actor may lawfully hold more than one of those powers, but the public record must still show the allocation before observers can assume it.

The court layer supplied the receivership and Board-restoration context. The receiver layer appointed the committee and approved process rules. The committee layer held collective election functions. AFRINIC’s finance layer classified and published the costs. An external auditor later addressed annual statements. Members were then asked to consider those statements. Each layer is real; none automatically supplies a document missing from another.

This is why a court-backed process is not a complete answer to a transaction question. A judicial or receiver mandate may establish that an election should be organised, and urgency may make professional assistance reasonable. It still does not state who could bind AFRINIC to USD 157,699, USD 55,484 or USD 44,605; who chose the provider attached to each line; who agreed the price basis; who accepted the work; or who reviewed and authorised any invoice or payment. The issue is not whether the recovery process had public significance. It is whether significance can substitute for documentary authority. It cannot.

Procurement and the formation of price

Nothing in the sealed record permits a general verdict on AFRINIC procurement, and this article does not offer one. The relevant question is transaction-specific: what selection and fee-review method applied to these three named engagements?

AFRINIC said the appointees were chosen for independence, legal background and election-process experience. That is a plausible specialist-selection rationale. In a compressed recovery process, an open tender might have been impractical or undesirable. A tailored invitation, urgency exception, market check, professional recommendation or negotiated team arrangement could all be defensible, depending on authority and circumstances.

The checked public set contains none of the records needed to determine which, if any, of those routes was used. There is no located invitation list, quotation, benchmark, exception memorandum, conflict waiver, purchase order or engagement approval. There is also no published independent fee review. Consequently, neither competition nor the absence of competition can be established; neither price reasonableness nor excess can be established.

The unequal amounts make the need for denominators more visible, but not more accusatory. The USD 102,215 difference between 3 Hare Court and Queens Court Chambers, the USD 113,094 difference between 3 Hare Court and Georges Penny Chambers, and the USD 10,879 difference between the latter two are arithmetically exact. They are not comparable prices unless the work, seniority, time, currency, taxes, disbursements and contractual structures are comparable. The public record supplies none of that equivalence.

A narrow accountability schedule could answer the question without disclosing substantive advice. It could identify the authorised client and supplier, engagement date, selection basis, non-confidential scope, fee model, cap, invoiced or accrued status, approving officer and independent review step. If urgency or specialist expertise displaced a broader process, the schedule could name the exception and approver. The present absence of such a schedule is a disclosure gap, not proof that no control existed.

Privilege should be neither dismissed nor inflated

Barristers performing candidate and election-process work may encounter confidential material, and some substantive communications may attract legal professional privilege depending on the client, purpose and applicable law. It would be irresponsible to demand publication of protected advice merely to satisfy curiosity. Candidate information and internal deliberation may also warrant careful treatment even where privilege is not the basis.

But privilege is a legal protection held by an entitled client, not a general word that resolves all documentary questions. The checked public record does not identify the client or privilege holder for these engagements. It contains no privilege log, confidentiality basis or redaction schedule explaining what has been withheld and why. The absence of such a record does not prove that privilege is invalid; it means the boundary is not publicly mapped.

Substantive advice can often be separated conceptually from control metadata. The names of contracting parties, appointment dates, broad non-confidential workstreams, fee structures, approval paths, aggregate billed status and payment classification do not automatically reveal advice. Some details may still require redaction, but a tailored schedule can state the ground and preserve the protected substance. Treating privilege as absolute would make basic authority and payment accountability impossible. Treating it as irrelevant would risk exposing legitimate confidential work. The appropriate demand is a reasoned boundary.

This matters directly to the “Georges Penny Chambers” discrepancy. A correction or payee-reconciliation statement need not disclose legal advice. Nor would a statement of whether the reported figure was paid, accrued, provided or reimbursed necessarily reveal the content of counsel’s work. Confidentiality may constrain the form of disclosure; it does not remove the need to identify which control questions remain capable of answer.

Conflict claims require attribution and process

AFRINIC’s FAQ said the committee appointees were chosen for independence, legal background and experience with election processes. That statement establishes AFRINIC’s rationale. It does not independently prove that every potential conflict was identified or resolved.

ICANN later publicly claimed that two Nomination Committee members had previously provided an opinion supporting registration of the same company as an AFRINIC member. ICANN called for reconstitution and remedial disclosure, and its page was later updated after a receiver request. This is a first-order governance fact because an external institution raised a concern about a committee controlling candidate and process functions. It remains an attributed claim. The sealed record does not permit this article to identify the individuals, declare that a conflict existed, or treat the allegation as adjudicated.

The missing public control record is more important than speculation about personalities. No checked declaration register, prior-engagement disclosure, recusal record, waiver, remediation decision or legal analysis closes the issue. A conflict process could have found no disqualifying interest, accepted a disclosure, imposed a recusal or reorganised a workstream. Or the factual premise could have been disputed. The public set does not establish which path, if any, occurred.

The financial connection is similarly bounded. A concern about committee independence does not prove taint in any one of the three cost lines. Without a mapping between people, duties, engagements and reported amounts, it is not even safe to attach the concern to a particular line. The proper accountability question is whether potential conflicts were declared, assessed by an authorised decision-maker, and managed in a documented way before work and costs were approved.

Invoice, payment and accounting language

AFRINIC reported the figures as costs. That term should be preserved. The checked record does not show whether each amount represented a cash payment, an accrual, a provision, a reimbursement, a tax-inclusive total or a translation from another currency. It contains no invoice, credit note, ledger entry, remittance, bank mandate, payment date or beneficiary confirmation.

This is not semantic caution for its own sake. The control questions differ by status. An accrued amount may still be disputed or awaiting payment. A reimbursed amount may involve a different supplier chain from a direct fee. A foreign-currency amount requires an exchange date and policy to explain the USD figure. Disbursements and tax can alter the relationship between professional time and the published total. Without the accounting basis, “paid” would imply a cash event the table does not prove.

Invoice review is also separate from publication. A finance team can classify an amount after receiving it, but classification does not show who certified that the work matched scope, whether a fee was checked against agreed terms, whether conflicts were cleared, or whether the reviewer was independent of the engagement decision. No checked record identifies an invoice approver or supplies a review note for any of the lines.

The same separation applies to audit. An annual auditor may sample, aggregate or test transactions according to financial-statement materiality. The public existence of an audit opinion does not reveal whether all three items were vouched, whether payees were confirmed, or whether authority and procurement were tested. If the auditor specifically reconciled these lines, a statement of that scope would improve assurance. In its absence, the safe description remains unaudited provider detail alongside a distinct later annual audit.

The annulment does not answer the value question

The June election process was annulled on 26 June 2025 after the receiver stated that concerns had arisen about powers of attorney and related voter documents. That event is consequential, but it must not be made to carry a causation claim it does not contain. The stated concerns were about voter documents. The record does not establish that the Nomination Committee’s work caused the annulment or that one of the three reported lines financed the disputed activity.

Nor does annulment prove that all work performed before it had zero value. Guidelines may have been drafted, candidates assessed, lists finalised and polls supervised even though the process did not produce a durable Board outcome. Professional work can be competently performed within a process that later fails for a different reason. Conversely, effort expended is not by itself proof of value. Value depends on authorised objectives, accepted deliverables, quality, usefulness and price—all still undocumented at line level.

The outcome therefore sharpens, but does not decide, the accountability question. When a process is later annulled, members have a stronger practical reason to understand which deliverables were completed, retained or reusable. They do not gain a factual basis to declare the advisers improper. A deliverables and acceptance record would allow the two issues to be separated: what work was completed, and what institutional outcome followed.

The strongest case for AFRINIC’s arrangement

The strongest contrary case deserves to be stated without caricature. AFRINIC was trying to reconstitute a missing Board under receivership and within a compressed timetable. An election with contested governance significance required rules, candidate scrutiny and supervision. A senior KC as chair, supported by three barristers associated with different chambers, could have offered parallel capacity, specialist legal judgment, resilience against one adviser becoming unavailable, and a visible degree of separation from ordinary AFRINIC management.

Different professionals may also have carried very different burdens. The largest line might reflect substantially more hours, a broader workstream, junior or administrative support, disbursements, or work undertaken for more than one committee member. The smaller lines might reflect narrower assignments. Confidential candidate and legal material could legitimately limit publication of detailed work product. Urgency could justify a specialist-selection method that would look different from routine purchasing.

The appointment communiqué at least made the names and collective duties public rather than concealing that external professionals were involved.

The later annulment does not defeat this case. It was attributed to concerns about voter documentation, not to a public finding that the committee’s professional work was incompetent or improper. Work undertaken before annulment may have been necessary and properly performed. A later annual audit and AFRINIC’s financial disclosure also added a degree of ex post visibility that would not exist if the costs had remained entirely unreported.

Every part of that case is plausible. None closes the documentary gaps. Parallel capacity is testable through a responsibility matrix. Specialist selection is testable through an approval and exception record. Unequal workload is testable through non-confidential scope, hours or deliverables. Privilege is testable through a log that separates protected substance from control metadata. Ex post disclosure is more useful when labels reconcile to suppliers and status. The fair conclusion is not that multiple advisers were unnecessary; it is that members cannot presently test the justification AFRINIC might reasonably possess.

The bookkeeper boundary and the people who bear the downside

Heng Lu’s controlling doctrine fixes the institutional frame. AFRINIC is a bookkeeper and coordinator for number-resource records. That necessary technical position does not confer sovereignty, taxation power, ownership power or punishment power. It cannot turn an internal classification into public-law authority, and it cannot make expenditure unreviewable merely because continuity or “the community” is invoked.

The doctrine is particularly relevant where control and liability diverge. The court, receiver, committee, professional advisers, AFRINIC finance function, auditor and members occupied different positions in the process. The members and dependent network operators bear the economic and operational downside of weak governance around the registry surface, yet the checked record does not show that they chose these engagements or approved individual invoices.

Institutional agents may make necessary decisions, but their authority must be specific enough to trace and their liability incentives must not be allowed to disappear into collective language.

NRS supplies a first-class accountability position at the member layer. Its 2026 member alert argued for transaction-specific authority, engagements, invoices, time records, approvals and payment evidence, and distinguished financial audit from legality, authority and reasonableness. That position is not evidence that NRS audited these three lines or proved a defect in them. It is a standard against which members can ask whether later account presentation gives them enough information for an informed decision.

LARUS supplies a first-class operator-risk perspective: governance failure at the registry layer can transmit continuity and asset-control risk to networks that depend on the ledger and cannot cheaply substitute another coordination surface. It does not supply a contract, invoice or fact about these three costs. Its relevance is to who absorbs downside when control is opaque, not to whether a particular bill was valid.

BTW’s first-class role here is a corpus boundary. Its existing coverage places adjacent aggregate and policy questions outside this article, leaving the three-line responsibility, naming and documentary-control map as the distinct object. BTW is not used as independent verification of the transactions.

Together, these sources support a disciplined consequence: USD 257,788 is material enough to warrant narrow transaction documentation, but no evidence in this package supports a claim that one member funded a particular cost, an operator suffered a quantified service loss, or the sum was excessive. Accountability is strongest when it asks for the missing link without pretending to know the answer.

What a complete public control record would show

A complete record need not publish privileged advice or sensitive candidate material. It would identify, for each reported line, the authorised client, contracting party and legal payee; the date and instrument of engagement; the approver and delegated authority; the broad workstream and intended deliverables; the procurement or specialist-selection route; the fee model, cap, currency and treatment of taxes and disbursements; the conflict-declaration and remediation status; the invoice reviewer; and whether the figure was invoiced, accrued, paid, provided or reimbursed.

At committee level, it would add a responsibility map covering the chair and three assistants. That map would explain how guideline design, candidate criteria, interviews, list finalisation, poll supervision and related work were allocated or shared. It would answer why the public personnel record contains four appointees while the finance table contains three chamber or person-linked lines, without assuming either omission or bundling.

For the USD 44,605 item, it would reconcile “Georges Penny Chambers” with George Hayden Penny at 187 Chambers. If the finance wording was shorthand or a transcription mistake, AFRINIC could state that and identify the correct legal counterparty. If it reflects another arrangement, the record could explain that relationship. The purpose is not to expose bank details or privileged work; it is to make the published supplier identity intelligible.

The record would also state the privilege boundary. A schedule could identify categories withheld, the privilege or confidentiality basis, the holder and the non-substantive metadata released. For conflict management, it could disclose that declarations were received, who assessed them and whether any recusal, waiver or change in duties occurred, without naming individuals as conflicted where no finding exists.

Finally, it would separate assurance stages. AFRINIC’s unaudited disclosure would remain what it is. Any transaction-specific reconciliation by management or the external auditor could be described within its actual scope. Any later member decision could be reported with its result and stated legal effect, rather than being assumed to cure or ratify earlier acts.

Until such records appear, uncertainty should remain explicit. The public does not know whether there was one engagement structure or several; who the legal payee was for each line; how work was divided; how the amounts were formed; what selection route applied; what was privileged; how conflicts were managed; who reviewed any invoice; whether the amounts were paid or accrued; whether the annual audit tested them individually; or what measurable result belonged to each line.

That uncertainty supports scrutiny, not a verdict. The most defensible conclusion is exact: AFRINIC’s own publication closes three labels, three amounts and the total of USD 257,788. Its appointment record closes four names and a collective mandate. It does not yet close the chain from appointment to engagement, from work to fee, from invoice to approval, or from accounting entry to payment.