Summary

  • Resolution 201304.168 adopted Delegation of Authority 002-2013 in line with a new organisation structure and added the COO office as a signatory of AFRINIC’s bank account. That was a meaningful change to who sat inside the organisation’s internal banking-execution perimeter, not a grant of sovereign, regulatory, police, punitive, confiscatory or adjudicative power.
  • Resolution 201304.169 performed a separate act: it named Anne Rachel Inné as the COO incumbent within a three-person principal signatory slate, alongside Adiel Ayodélé Akplogan as CEO and Patrisse Déesse as Director Finance and Accounting. The separation makes office and person visible as distinct objects, but it does not show how later changes of incumbent would be handled.
  • Resolutions 201304.170-172 complete only the public outline. They identify two alternates under a stated absence condition, preserve Board approval for transactions above USD 100,000 and instruct that the relevant policies and resolutions be furnished to AFRINIC’s bankers. They do not publish the ordinary combination matrix, lower thresholds, threshold scope, alternate activation mechanics, bank acceptance, effective date or express revocation of earlier authority.
  • The strongest benign interpretation is that the public register was a concise notice sitting above a complete confidential delegation and a correctly implemented bank mandate. That may be entirely true. It still leaves a bounded disclosure opportunity: AFRINIC could publish version, role categories, categorical combinations, threshold scope, activation and revocation status without exposing account numbers, credentials, signatures or attack-enabling procedures.

A role entered the execution perimeter

The precise starting point is unusually important. In April 2013, AFRINIC’s Board adopted a new Delegation of Authority, numbered 002-2013, in line with a new organisation structure. Resolution 201304.168 says the change added the Chief Operating Officer as a signatory of AFRINIC’s bank account. The resolution speaks in the language of an office. It does not name the person occupying that office in the sentence that makes the amendment.

That distinction prevents two opposite errors. One would be to treat the sentence as a trivial change of job title, as though nothing operational followed from placing another executive role within the banking perimeter. The other would be to treat the sentence as proof of a complete and immediately effective personal mandate, including powers, combinations and bank activation that the published wording does not show. The sound reading is between those extremes. The Board changed the architecture of delegated execution. That change was real. The public sentence alone does not disclose every rule that made it operational.

For a private coordinator funded to maintain a technical and organisational function, banking authority is consequential. It determines which offices may participate in instructions concerning the organisation’s own money and how executive continuity can be maintained when a principal officer is unavailable. Adding the COO can increase operational capacity and reduce dependence on a narrower executive grouping. It can also widen the class of people positioned to participate in execution.

Whether that widening is well controlled depends on the combination rule, the distinction between initiation and approval, the applicable thresholds, the alternate arrangements and the revocation process. The resolution announces the office-level amendment but leaves those supporting mechanics in the referenced delegation.

This is meaningful internal authority, but it is not public authority. AFRINIC is a private technical bookkeeper and coordinator. It is not a sovereign, government, regulator, police force, prosecutor, punishment authority, confiscator or court. A mandate concerning AFRINIC’s own bank account cannot turn a coordination body into any of those institutions. It does not confer power over the Internet, routing, operators, networks or number-resource property. The treasury serves the coordinator’s operational reality in the same way that its technical ledger does: as infrastructure for a limited function, not a source of political supremacy.

The public significance of Resolution 201304.168 therefore lies in control design. A new organisational structure was being translated into an authority map. The Board chose to put the COO role into that map. The questions that follow are not whether the COO acquired governmental power, nor whether a transaction went wrong. They are narrower and more answerable: Which office was authorised? Which individual was named? In what combinations could authorised people act? What approval boundary remained with the Board? When did the bank make the change effective? What old authority ceased, and when?

Within roughly the first movement of the public record, the analytical frame is already complete. An internal execution-control change matters because it affects custody and continuity inside a member-funded private organisation. Its importance does not require exaggeration. Indeed, exaggerating it would obscure the real governance problem. The proper demand is a thin, explicit, role-aware, auditable and reversible authority chain proportionate to AFRINIC’s private coordination function.

Resolution 201304.168 authorised an office; Resolution 201304.169 named people

The immediately adjacent resolution supplies the next link, but it must not be collapsed into the first. Resolution 201304.169 separately names three people and their offices as bank-account signatories under the Delegation of Authority: Adiel Ayodélé Akplogan as CEO, Patrisse Déesse as Director Finance and Accounting, and Anne Rachel Inné as COO. The COO incumbent is therefore visible in the public sequence, but the naming act belongs to Resolution 201304.169 rather than to Resolution 201304.168.

That two-step structure is valuable evidence. The first decision concerns institutional design: the COO office enters the delegation. The second concerns personal designation: the named incumbent appears in the signatory slate alongside the other named officeholders. An office persists as an organisational position; a person’s tenure in it can begin and end. A robust authority record has to know both. If it records only the office, a bank may lack the verified human identity it needs to apply the mandate.

If it records only a name, later readers may not know whether the authority belonged to a continuing role, a specific personal appointment or both.

The published separation does not answer the succession question. It does not say whether every future COO would become authorised simply by taking office, whether the Board would have to adopt another personal designation, whether the bank would require a fresh mandate, or whether some combination of those steps applied. No answer should be invented. The fact that the Board separately adopted a role-level amendment and a named slate suggests that the objects were treated separately in this sequence; it does not establish a permanent rule for every later personnel transition.

The distinction also blocks an unwarranted inference about unilateral power. Naming someone as a signatory does not, by itself, say that the person could issue a valid instruction alone. A signatory may be one participant in a multi-person combination. The mandate may assign different roles to initiation, checking, approval or execution. It may vary by account or amount. None of those details appears in the public sentence. The missing table matters precisely because the word “signatory” identifies eligibility to participate without necessarily disclosing the conditions under which a signature or instruction becomes effective.

Nor does the naming establish that any named person ever initiated, approved or executed a particular transaction. The closed public record connects no person in the sequence to any payment, withdrawal, transfer, loss or movement of funds. It identifies authority design and designation. It does not document use. Any narrative of personal conduct, good or bad, would move beyond the evidence.

Keeping the office and person distinct is not bureaucratic fussiness. It is the basis of reversible control. When an incumbent leaves, the organisational role may remain within the approved design while that person’s bank authority must end. When a new incumbent arrives, the role may already be recognised in the delegation while personal identification, due diligence and bank activation still have to occur. A public change record can acknowledge that sequence without publishing private identity documents or financial-security material.

The point of examining Resolution 201304.169 here is limited. Its full three-person designation is a separate subject. In this investigation, it illuminates the office-to-incumbent bridge created around the COO amendment. The earlier 2011 authorised-signatory slate likewise provides a baseline but is not being reconstructed. By maintaining those boundaries, the 2013 office-level change can be understood on its own terms rather than becoming a general history of AFRINIC’s banking arrangements.

The adjacent resolutions expose the outline of a control chain

Resolutions 201304.170, 201304.171 and 201304.172 follow the office amendment and named designations. Read together, they show that the Board did more than add a title to a list. It considered alternate signatories, retained an exception requiring Board approval above a stated amount, and directed that the relevant materials be sent to the bankers. Those are recognisable layers in an authority chain. Each layer also has a boundary that the public wording does not cross.

Resolution 201304.170 addresses absence. It states that, in the absence of two of the three appointed signatories, Dr Krishna Seeburn, as Board Representative in Mauritius, and Christian Fanchette, as Director Human Resource and Administration, would also serve as alternate signatories under the Delegation of Authority. The resolution therefore provides a continuity mechanism tied to a condition. It does not show the ordinary signing combination, explain how absence would be established, specify whether one or both alternates became active, or say whether their activation applied to one transaction or continued across a period.

Those are not small semantic gaps. If the normal rule required two participants, the absence of two principals could leave only one principal available, but the public sentence does not say what combination followed. If a different matrix applied, another configuration might be intended. The words “would also serve” establish alternate eligibility under the stated condition; they do not establish permanent activation. Treating both alternates as continuously active would overread the text. Treating activation as automatic would do the same.

Resolution 201304.171 identifies a high-level boundary. It authorises the named signatory persons to act under the AFRINIC authorisation policy described in the Delegation of Authority table, subject to an exception: a transaction in excess of USD 100,000, or the equivalent in another currency, had to be approved by the Board. The sentence proves that the unseen table governed ordinary authority and that high-value activity crossed a Board-approval boundary. It does not reproduce the table.

The distinction between approval and signing is central. Board approval above a threshold is not necessarily a rule about how many signatures a bank instruction requires. Approval can be a prior corporate decision; execution can be the later act of sending or validating an instruction under a combination matrix. A transaction could require Board approval and still need whatever executive signatures the delegation prescribed. Conversely, the existence of authorised signatories does not make the Board exception irrelevant. The public page shows both layers but not the exact interface between them.

The amount itself must remain bounded context here. An earlier 2011 resolution and the analysis devoted to it own the detailed investigation of the USD 100,000 boundary. The 2013 sequence matters because it shows that the exception accompanied the new delegation and named signatories. It does not disclose whether multiple related instructions were aggregated, which accounts were covered, when a foreign-currency equivalent was measured, or how transfers between AFRINIC accounts were treated. Those questions remain unanswered; they cannot be converted into asserted rules.

Resolution 201304.172 supplies the external-implementation link in outline. It says AFRINIC’s bankers were to be furnished with copies of the relevant policies and resolutions, including the resolution. That instruction matters because a corporate decision inside a Board record and an operative control inside a bank’s systems are not the same event. Furnishing the documents is the bridge by which the external institution could update its mandate and controls.

But an instruction to furnish is not proof of delivery, acceptance or activation on the same date. The public register does not publish a transmission date, an acceptance receipt, completion of any required verification, a bank-control update or confirmation that the mandate became operative. This does not show that any of those steps failed to occur. It shows only that the public sequence stops before confirming them.

Together, the five resolutions form a coherent visible outline: adopt the delegation and add the COO role; name the principal people; provide alternates for a defined absence condition; preserve Board approval above a high-value boundary; send the governing materials to the bankers. The outline is stronger than a solitary sentence. It demonstrates a considered sequence of internal actions. It still does not supply the ordinary authority matrix or the dated implementation and revocation bridge.

The unseen table holds the operational grammar

Delegation of Authority 002-2013 is referenced, but its text and table are not reproduced on the public Board page. That absence establishes the limit of what can be concluded. The table may have been complete, carefully designed and appropriately confidential. It may have specified every combination and lower threshold needed for safe operation. The public source does not allow a reader to verify those propositions.

An authority table is where a broad role designation becomes an operating rule. It can distinguish who may initiate an instruction from who may approve it. It can prescribe whether members of different role categories must act together. It can set different boundaries for different types of account or instruction. It can state lower delegated amounts below the high-value Board exception. Without the table, the public record tells us who was placed inside the signatory perimeter but not the normal grammar of valid action.

The missing combination is the most immediate problem. One cannot infer that any one named person could act alone. One cannot infer that any pair was required. One cannot infer that the COO had to act with the finance director, the CEO or an alternate. Each possible reconstruction adds a rule that is not published. The disciplined conclusion is that the combination remains unproved in the closed public record.

The missing rights distinction follows. “Signatory” does not disclose whether the relevant person could initiate, approve, view, confirm or execute, nor whether those capacities varied across accounts. Banking controls often separate functions conceptually, but the public evidence here does not reveal the adopted separation. The analysis therefore cannot credit the COO with a particular functional right beyond inclusion as a bank-account signatory under the delegation.

Lower thresholds are also invisible. Resolution 201304.171 exposes the Board exception above USD 100,000 but points ordinary operation back to the table. The register does not show whether lesser values were divided into bands, whether different offices had different ceilings, or whether the combination varied with amount. It also does not settle transaction scope, aggregation or currency-measurement questions. A public high-value exception is a ceiling on delegated approval at one layer; it is not a complete picture of controls beneath it.

The account scope is not established either. The resolution uses the formulation of AFRINIC’s bank account, while the missing table may have held more precise operational distinctions. It would be unsafe to infer an account-by-account allocation or universal coverage. Similarly, nothing proves how inter-account transfers were classified. Precision requires leaving those matters open.

This evidentiary restraint is not a verdict against AFRINIC. Public summaries are often shorter than the instruments they summarise. The proper finding is about auditability: outsiders can see that a change was adopted and can see several supporting resolutions, but cannot reconstruct the ordinary control design from the published material alone. The finding remains true whether the confidential system was excellent, adequate or flawed.

The governance question is therefore not “Where is every sensitive banking instruction?” Publishing those details could be irresponsible. The question is “Which non-sensitive control properties could be confirmed?” A categorical matrix could say, for example, that principal roles operate in a defined multi-role combination without naming credentials or describing the bank’s security process. It could state whether alternates require documented activation, whether the threshold applies per instruction or under another defined scope, and when the bank confirmed the revised mandate. Such disclosure would expose control shape, not attack surface.

Activation is an event, not an assumption

Board adoption, notification to bankers, bank acceptance and control activation are distinct points in the chain. The public register proves the first and records an instruction directed toward the second. It does not publish dates or receipts for the remaining steps. That matters whenever authority is changed, because a period can exist between institutional decision and external implementation even when every participant acts properly.

Resolution 201304.168 belongs safely to April 2013. The Board register groups it with the April resolutions. A later June entry approved minutes of meetings held on 24 April and 28 May, but the public page does not assign Resolution 201304.168 to a particular meeting day. The evidence therefore supports the month, not an exact date. The effective date is likewise not stated.

An approval date and an effective date can coincide, but they need not. The delegation itself may have contained an effectiveness clause. The bankers may have required documentation or verification before changing the mandate. The organisation may have planned a transition interval. None of those possibilities can be confirmed from the public page. The correct formulation is that the Board adopted the new delegation in April and instructed that relevant materials be furnished; the date when bank controls became active remains unproved.

Bank acceptance deserves separate treatment because it is the point at which an internal authority decision becomes enforceable within the external banking arrangement. A notice can be sent but require clarification. A bank can receive resolutions but still need identification or its own mandate forms. Again, these are possible links in the logic of implementation, not claims about what happened in this case. No failure can be inferred from the absence of a public receipt.

A compact implementation record would resolve the uncertainty without revealing sensitive details. It could state the delegation version, date of Board approval, date delivered, date accepted and date activated. It could confirm that required verification was completed without publishing private documents. That would let members distinguish a resolution that existed on paper from a control that was operative at the bank.

The value of such a receipt is longitudinal. Years later, readers should not have to assume that the publication date, meeting date, delivery date and activation date were identical. A simple dated chain protects both the institution and its members from speculation. It makes the transition reviewable while avoiding any disclosure of account identifiers, signatures, credentials or transaction procedures.

Revocation closes the change

Adding new authority is only half of a safe transition. The other half is identifying what previous authority was superseded and when any obsolete personal mandate ended. The April 2013 sequence does not expressly state which earlier delegation, named mandate or signatory authority was revoked or replaced. It does not publish a transition interval.

The earlier 2011 resolutions establish that prior signatory and threshold arrangements existed, but their full investigation belongs elsewhere. Here, they serve one narrow purpose: they show why a supersession link would be useful. When a Board adopts Delegation 002-2013, the version number and label suggest a defined instrument, yet the public register does not identify the exact predecessor that ceased to govern or the date of cessation.

It would be wrong to infer that adoption automatically revoked every earlier mandate. A new internal policy may supersede an earlier one as a matter of corporate documentation, while personal authorities at a bank may require separate removal. Or the new delegation and bank documents may have handled both at once. The public source does not decide between those possibilities.

The office-person distinction returns here with practical force. If the COO role stays authorised after an incumbent departs, the office-level design can remain valid while that individual’s personal mandate must be deactivated. If a role is removed from the organisation structure, both the design and the incumbent’s authority may require change. A reversible system needs an express record for each layer rather than relying on assumption.

Revocation also protects against a misleading archive. A reader who sees only a new slate may assume old names disappeared operationally at once. A reader who sees only a historical resolution may assume a named person retained authority indefinitely. Neither inference is justified. A public status line can say that superseded personal mandates were revoked on a specified date without exposing how the bank processed them.

The minimum useful chain would identify Delegation 002-2013 by version and approval date, identify the precisely superseded version, separate roles from named incumbents, show the effective date and confirm bank acceptance and revocation status. It should also assign periodic review ownership and retain a compact change or exception log. That is not a demand for transaction disclosure. It is a demand for the life cycle of authority itself.

The strongest benign reading

The most persuasive countercase should be stated in full. AFRINIC had adopted a new organisation structure and needed its banking arrangements to follow it. The Board used a coherent five-resolution sequence rather than a single vague instruction. It adopted a new delegation, named the principals, provided for alternates, retained Board approval above USD 100,000 and directed that the policies and resolutions be furnished to the bankers. The full delegation table may have contained all ordinary combinations, lower thresholds and activation rules. The bankers may have implemented the mandate correctly.

A concise public register was not necessarily intended to reproduce the operational instrument.

Operational security strengthens that case. Publicly exposing account numbers, credentials, specimen signatures, security procedures or detailed bank-control instructions could create risk without improving legitimate accountability. A member-funded coordinator does not owe the public a manual for reaching its accounts. It can reasonably keep sensitive material between authorised officers, its Board and the bank.

That benign reading is consistent with every established fact. Nothing in the public record proves a control failure. Nothing shows that any unauthorised person acted, any threshold was breached, any transaction was mishandled or any money was lost. No negligence, corruption, fraud, theft, misappropriation, illegality, bad faith or personal fault follows from the disclosure gaps. The named people should not be turned into characters in an invented transaction narrative.

But the countercase does not eliminate the auditability question. A register can remain concise while confirming non-sensitive properties. It can identify the delegation version and predecessor. It can list role categories and named incumbents separately. It can describe ordinary combinations at a categorical level, such as whether multiple roles must participate, without revealing the bank’s technical process. It can define the scope of the Board threshold, the condition and duration of alternate activation, and the dates of acceptance, activation and revocation.

This bounded disclosure would support the benign reading rather than undermine it. It would demonstrate that the internal instrument and bank implementation formed a closed chain. It would prevent later readers from mistaking silence for unilateral authority, continuous alternate status or automatic revocation. It would also reduce pressure to disclose sensitive documents, because a compact control notice could answer legitimate governance questions without exposing operational detail.

The narrow finding, then, is neither that AFRINIC disclosed nothing nor that the control system was defective. The Board register exposes a meaningful and apparently structured change, but it stops before the ordinary combination matrix and implementation life cycle. That limit constrains what can be audited publicly. It does not establish what happened privately.

Private corporate authority must remain private in constitutional meaning

The institutional boundary is direct. AFRINIC is a private technical coordinator and bookkeeper. Its internal governance powers can be real within their proper domain: it can organise staff, approve policies, control its bank accounts and delegate execution within the organisation. Those capacities are necessary to perform a private coordinating function. They do not become sovereign powers because the institution performs important technical work.

The distinction matters especially when discussing “authority.” The word can describe permission inside a corporate control system, as it does here. It can also be used to imply governmental jurisdiction over people, property or infrastructure. Moving between those meanings without warning manufactures power the evidence does not support. Resolution 201304.168 concerns the first meaning only: an office was admitted to a private bank-signatory arrangement over AFRINIC’s own account.

No aspect of the decision makes AFRINIC a regulator of the Internet, a police body for operators, a prosecutor, a punishment authority, a confiscator of number resources or a court. It does not create adjudicative competence. It does not make internal financial execution analogous to public law. The importance of sound banking controls comes from stewardship of the coordinator’s resources, not from any supposed sovereignty.

This limitation is not an argument that internal governance is unimportant. The opposite is true. Because AFRINIC is a narrow coordinator rather than a state, its controls should be proportionate, intelligible and traceable. Officeholders exercise bounded functions in service of operational reality. The institution should be able to show where a delegated function begins, what conditions constrain it and how it ends.

The ledger and treasury are parallel in this limited sense. Both are tools serving actual network coordination and organisational operation. Neither should be used as a stage for institutional grandeur. A technical ledger does not turn its keeper into a sovereign over the underlying networks; a treasury mandate does not turn its signatories into public officials with coercive jurisdiction. In both settings, authority should be thin enough to audit and reversible enough to correct.

For members and operators, this provides the right scale of concern. They have a legitimate interest in whether a member-funded coordinator can execute payments reliably, maintain continuity during absence, respect Board approval boundaries and remove obsolete authority. They do not need access to credentials or individual transaction files to assess those control properties. Nor should concerns about internal controls be inflated into theories of public power over their networks or resources.

Institutional legitimacy grows from staying inside the mandate. AFRINIC can demonstrate competent stewardship through a clear delegation record while recognising that its broader function remains private coordination. The April 2013 sequence is best examined on precisely that basis: a genuine internal change, deserving exact accountability, confined to an organisation that cannot legislate, police, punish, confiscate or adjudicate the Internet.

What the public evidence proves—and what it cannot carry

The public evidence proves a bounded set of propositions. In April 2013, the Board adopted Delegation of Authority 002-2013 in line with a new organisation structure and added the COO role as an AFRINIC bank-account signatory. The adjacent resolution named Anne Rachel Inné as COO within a three-person principal slate that also named the CEO and Director Finance and Accounting. Another resolution identified two alternates under the condition described as the absence of two of the three appointed signatories. A further resolution retained Board approval for transactions above USD 100,000 or the equivalent in another currency.

The final adjacent resolution instructed that relevant policies and resolutions be furnished to the bankers.

The evidence does not prove the ordinary signing combination. It does not prove unilateral authority for any person, a required pair, a particular cross-role combination or an account-specific rule. It does not disclose lower thresholds or distinguish initiation, approval, viewing and execution rights. It does not show how absence was evidenced, who activated an alternate, whether one or both alternates served, how long activation lasted or how it ended.

It does not define whether the USD 100,000 exception applied per instruction, under an aggregation rule or across a specified account scope. It does not establish the foreign-exchange measurement time. It does not state how an inter-account transfer would be treated. Those details cannot be borrowed from the mere existence of a threshold.

The evidence also does not prove an exact April adoption day, immediate effectiveness, bank receipt, bank acceptance, completion of verification, system activation or implementation confirmation. It does not expressly identify the prior delegation or personal mandates that were revoked, nor when any revocation took effect.

Finally, it proves no specific use of authority. There is no established transaction, payment, withdrawal, transfer, account movement, threshold breach, loss, irregularity or control failure attached to the resolution sequence. There is no evidentiary basis for attributing misconduct or fault to AFRINIC or to any named person.

This line between proof and possibility is the centre of responsible institutional analysis. Missing documentation should not be converted into presumed failure. Institutional wording should not be converted into presumed implementation. A register entry proves what it records. The unseen instrument may contain more; the bank’s records may confirm more. Until those links are made public in an appropriate form, they remain outside the verifiable chain.

A proportionate remedy: publish the shape, protect the mechanism

The appropriate remedy is a compact authority-change notice, not wholesale disclosure of banking operations. It should identify the delegation by document number, version date and approval date. It should name the superseded version precisely. It should separate the continuing organisational roles from the people currently designated for those roles.

It should state ordinary signature combinations categorically. The disclosure need not reveal account numbers, credentials, specimen signatures, authentication methods or step-by-step procedures. It only needs to make clear whether valid execution requires more than one role, whether role categories must be mixed and whether the combination changes by authority band. This is enough to prevent an observer from confusing signatory eligibility with unilateral power.

It should define the alternate-control life cycle. The notice should state the trigger, what constitutes the relevant absence, who confirms the trigger, whether one or both alternates are activated, how long the activation lasts and who deactivates it. Again, no personal security material is needed. The objective is to distinguish conditional backup capacity from standing authority.

It should describe threshold scope without turning the notice into a transaction manual. The Board exception can be tied to a defined measurement unit and account scope. The record can say whether related instructions are considered together, when currency equivalents are measured and how internal transfers are classified. Those categories permit governance review without publishing any actual transaction.

It should close the implementation chain with dates. Approval, notification, bank acceptance and activation should be separate fields. The record should confirm that required verification occurred while keeping the underlying private materials confidential. It should state which earlier mandates were revoked and on what date. If a transition interval existed, its beginning and end should be visible.

Finally, the system should assign a periodic review owner and retain a compact exception and change log. A delegation can become stale when organisational structure or personnel changes. Periodic review makes the role-person map current. An exception log allows governance bodies to see whether the designed controls are being altered without requiring publication of transaction detail.

The result would be both more transparent and more secure. Members could verify that the authority chain was coherent. AFRINIC could protect operational banking information. Named officers would be less exposed to false inferences about powers they may not have held. Future reviewers would no longer need to guess whether a new resolution automatically displaced an old mandate or whether a conditional alternate remained active.

This is the correct scale of reform for a private coordinator. It improves stewardship without pretending that corporate administration is public law. It strengthens accountability without creating an attack surface. And it treats authority as a life cycle—from role design to personal designation, combination, approval, activation and revocation—rather than as a single word in a Board register.