Summary

  • On 25 June 2026, AFRINIC’s AGMM agenda gave ten minutes to presentation of the already approved 2026 budget. Its listed ordinary resolutions asked members to approve the 2022–2025 audited financial statements and authorise appointment and remuneration of the 2026 external auditor; none asked members to approve the 2026 budget.
  • That sequence does not by itself violate AFRINIC’s Bylaws. Articles 15.1 and 15.3 place the company’s business under Board direction and give the Board power to determine the budget and establish or vary expenditure ceilings. The accountability problem is therefore not the absence of a member power the Bylaws do not grant. It is the absence of a timely choice before large or exceptional costs become contractual, paid or operationally difficult to reverse.
  • The scale makes timing consequential. AFRINIC disclosed USD 3,289,408 in legal expenditure across 2022–2025 and USD 1,043,425 in costs for two 2025 Board elections, including USD 300,000 in Receiver fees. Historical accounts can show how costs were recorded, but they do not by themselves identify who authorised every engagement or payment, whether alternatives were available, or what outcome each commitment purchased.
  • A workable remedy would preserve Board and management execution while separating a thin core-service floor from discretionary and exceptional spending. Members should receive a forward commitment calendar, transaction-specific authority records and executable cost envelopes before the last practical cancellation or renewal date—not a referendum on every invoice and not a sovereign role for a private registry.

L3 — AFRINIC members saw the 2026 budget after approval

At 12:20 UTC on 25 June 2026, according to AFRINIC’s published agenda, the organisation’s annual general members’ meeting reached its budget item. Ten minutes were allotted to presentation of the 2026 budget. The formal decisions listed elsewhere on the agenda looked backward: ordinary resolutions concerned the audited financial statements for 2022, 2023, 2024 and 2025. Another resolution concerned authority to appoint and remunerate the external auditor for 2026. There was no listed ordinary resolution asking the members to approve the annual budget then being presented.

That difference is easy to lose in the ceremony of an annual meeting. A document can be circulated. Slides can be delivered. Members can ask questions, object, praise or criticise. None of that is meaningless. But information and control are not the same institutional act. Information tells a principal what a company plans or has done. Control exists only if the principal can choose among feasible alternatives while the material cost remains avoidable.

By the time a commitment has been signed, renewed, paid, staffed or built into operational dependencies, the opportunity to speak may remain while the opportunity to change the decision has narrowed or vanished.

AFRINIC’s own published budgeting process makes the sequence explicit. Budget work begins in September. Management, Finance and the Finance Committee prepare and review the plan. The Board approves it by year-end or in the first weeks of January. The approved budget is then posted within fourteen days and presented at the next AGMM. The precise date and text of the resolution approving the 2026 budget are not publicly established by the records considered here.

What is established is the structure: approval belongs earlier in the cycle, presentation to members comes later, and the operating year is already under way by the time of the annual meeting.

The Bylaws explain why. AFRINIC is a Mauritian private company limited by guarantee. Articles 15.1 and 15.3 put direction of the company’s business in the hands of the Board and empower the Board to determine a financial budget as well as establish or vary expenditure ceilings. The Bylaws also use a different control for fees: a distinct sixty-day consultation applies to fee changes. Revenue incidence and spending authority are therefore separate questions in AFRINIC’s corporate design.

The text cited here does not grant members an annual right to approve the budget, and it would be inaccurate to convert a criticism of timing into a claim that the June agenda breached a nonexistent member-budget-vote requirement.

The sharper criticism starts by accepting that legal allocation. A Board may possess the corporate power to decide a budget, yet members may still receive too little prospective control over the exceptional uses of the money they supply. Bylaw competence answers who may act for the company. It does not automatically answer whether the process gives fee-paying members enough warning, meaningful alternatives or a usable authority trail. A validly allocated power can still be exercised through an accountability design that is too retrospective for the economic consequences it creates.

The 2026 numbers show the size of that consequence. The approved budget projected USD 6,635,936 in revenue and USD 6,609,595 in operating expenses, leaving a projected operating surplus of USD 26,341. Fee revenue was projected at USD 6,375,936, about 96.1 per cent of total projected revenue. In other words, members were expected to finance nearly the entire operating envelope. The surplus figure is an operating comparison, not a claim about cash remaining after every investment: capital expenditure was presented separately and should not simply be added or subtracted without respecting that presentation.

Within operating expenditure, the budget allocated USD 2,901,811 to human resources and described staff cost as about 44 per cent of operating expense. It placed USD 460,000 in travel and USD 400,000 in meetings, together about 13 per cent of projected operating expense. It included USD 290,000 for legal fees, of which USD 250,000 was stated to be for external legal advisers. Separately, the budget displayed USD 1,265,875 in capital expenditure, including USD 650,000 for MyAFRINIC v2 and USD 573,200 for the Infrastructure Unit.

These are not interchangeable kinds of cost. Staff continuity has a different commitment clock from an airline reservation. A legal instruction has a different renewal and exit structure from a meeting venue. A capital system has procurement milestones, implementation dependencies and switching costs that do not resemble either. The USD 650,000 MyAFRINIC v2 item illustrates the point. AFRINIC described the existing version as five years beyond end of life and as a single point of failure. That description may make replacement urgent and essential.

It still does not reveal when a vendor was selected, what price alternatives existed, when cancellation stopped being practical, what service dependencies arose or which decision point remained open to members.

A conventional annual budget compresses all of these clocks into one twelve-month table. That table is useful for planning but weak as a map of member agency. If members see a line in June, the line does not tell them whether the corresponding contract was signed in January, whether the project can be reduced in July, whether a reservation can be cancelled, whether a retainer renews automatically, whether recruitment has already created obligations, or whether an external adviser has already been instructed to undertake work. A budget states an intended envelope.

It does not show the moment at which each material amount crosses from proposal into sunk or difficult-to-reverse cost.

Nor should the approved budget be mistaken for actual audited expenditure. It is a prospective plan. The amount ultimately spent on legal work, travel, meetings, staff or capital systems may be lower or higher, and classifications can change. This limitation cuts both ways. A member cannot infer that every dollar was already committed by 25 June. But neither can AFRINIC infer that presentation in June gave members a practical choice over amounts that may already have entered contracts or operational plans. Without a commitment record, the degree of reversibility is unknown.

The AGMM’s account resolutions occupied a different time horizon. Audited financial statements concern recorded transactions and balances from completed periods. Member approval of such statements may have consequences under the applicable corporate process, but it is not proof that every underlying engagement was properly authorised, sensibly procured or effective. Accounting recognition, corporate authority, commercial reasonableness and legal outcome are different tests. A clean statement of what was paid does not identify the person or instrument that authorised each payment.

A correct total does not establish whether the service represented value. A vote on accounts cannot casually be described as legal ratification, waiver or cure unless a competent instrument establishes that effect.

This distinction matters because institutional language can blur chronology. “The members approved the accounts” sounds broader than it is. It can be heard as “the members approved the decisions reflected in the accounts,” even when the listed resolution concerned the financial statements rather than each transaction. “The budget was presented to members” can likewise be heard as “members had a say in the budget,” even when the Board had already approved it and no budget resolution was on the agenda.

Accurate governance language must keep the verbs attached to their objects: the Board approved the budget; members received its presentation; members were asked to decide on historical statements and the appointment of an auditor.

The absence of a listed budget resolution does not establish that no member discussed the 2026 plan or tried to influence it. The agenda alone cannot show every exchange in the meeting. Members might have questioned priorities, requested disclosure or proposed later action. Those possibilities are compatible with the documented sequence. They do not change the central timing question: what expenditure was still avoidable when the presentation began, and what executable alternative could a member decision then select?

That is the control surface this analysis examines. It runs from management’s proposal through Finance and Finance Committee review to Board approval, any variation of expenditure ceilings, the making of contracts or operating commitments, publication of the approved plan, current-year monitoring and eventual presentation or historical account vote. At each stage, a different actor may possess information or authority. The purpose of mapping the chain is not to turn the members into managers. It is to identify the last date on which a decision is still economically real.

AFRINIC’s role makes restraint especially important. It performs a useful technical function: maintaining a reliable register of Internet number resources, coordinating uniqueness and delivering bounded registration services. It is a private bookkeeper and coordinator, not Africa itself and not a public authority. It has no sovereign, regulatory, police, prosecutorial, punitive, confiscatory or public-law adjudicative power. Its meetings and elections allocate responsibility inside a company. They do not turn private corporate consent into continental consent or confer authority over states and non-consenting operators.

That narrow status does not make the budget trivial. On the contrary, control of an indispensable administrative record can produce practical leverage even without public-law authority. Operators depend on accurate recognition and continuity. Exit may be costly or constrained by the function the registry performs. When those who depend on the record also finance nearly all of its operating plan, delayed financial choice can deepen lock-in: they remain able to speak, but they cannot cheaply undo a commitment made in the institution on which they rely. The answer is more disciplined private accountability, not a larger claim of registry power.

The June 2026 meeting therefore revealed a structural gap rather than a simple procedural error. The company’s corporate rules can authorise the Board to approve the annual plan. The published process can make that plan visible. The auditor and members can later examine historical statements. Yet none of those elements, alone, tells a member which large costs were still open to choice before they hardened. A presentation is a disclosure event. An audit is a retrospective assurance mechanism. Neither is a prospective commitment calendar.