Summary

  • ISIF Asia is substantial enough to shape the Internet-development field it serves. APNIC reported that by January 2025 the programme had supported 166 grants and 43 awards, allocating more than USD 10.1 million across 32 economies. The 2023 round awarded about USD 2.284 million to 24 projects in 17 economies; the 2024 round awarded USD 1.675 million to 21 projects in 17 economies. These totals establish reach and financial importance, not impact or influence.
  • The APNIC Executive Council initiated the APNIC Foundation in 2014, the Foundation was established in 2016 and became operational in 2017, and administration of ISIF Asia moved from APNIC to the Foundation. The arrangement can attract money beyond membership fees and add specialist capacity. It also means that members need a clear map of appointments, shared services, related-party transactions, donor restrictions and decisions across APNIC, the Foundation and the Asia Pacific Internet Development Trust.
  • A grant does not prove support was purchased, and a recipient's later praise, conference participation or election activity does not prove causation. The relevant risk is structural: repeat funding, invitations, visibility, references and committee access can make it costly for recipients to criticize the institution or can make the institution mistake a funded network for independent consent.
  • Selection should be insulated from institutional politics. Published criteria, scored decisions, reviewer identities and conflicts, reasons for exceptions, unsuccessful-applicant distributions and a complaint route are more important than a list of winning projects. Invitation-only funds require a separate register because discretion begins before an application exists.
  • Impact evaluation must follow the funded public good beyond delivery. Infrastructure grants need operation, adoption, security, cost and continuity evidence; training needs retained competence and use; research needs methods, data and influence that can be examined; inclusion projects need beneficiary denominators and safeguards. Testimonials and recipient visibility cannot substitute for counterfactual evidence.
  • Grant recipients remain free citizens and organisations. They should not lose ordinary member rights or be silenced. The proportionate safeguard is disclosure and recusal on decisions directly affecting grants, a cooling-off period before serving on grant-selection or programme-oversight bodies, and separate reporting of later institutional roles so financial dependence is not confused with representative authority.

The public good and the political side effect can coexist

The best argument for an Internet-development grant is practical. A rural provider may need capital to deploy IPv6 when customers will not pay a premium for it. A university team may have a useful routing-security idea that cannot attract venture finance. An exchange point may need modest equipment and technical assistance whose regional benefit exceeds the recipient's private return. A disability-access project can solve a problem ignored by mainstream platforms. A small grant can move quickly where procurement, commercial credit and government programmes do not.

The political side effect begins with the same transaction. An organisation applies to a body that defines the categories, judges the proposal, pays the money, publicizes the winner and may invite the recipient to conferences. Success can help the recipient hire staff, complete a project, acquire references and enter regional professional networks. A second application may depend on a record of cooperation. Even without an explicit request for loyalty, criticism can begin to feel costly.

That possibility does not make the grant corrupt. Gratitude is a normal human response, and many recipients will criticize a funder when necessary. Some will have opposed it before and after an award. Others will become stronger, more independent actors because the grant diversified their income. The governance mistake is to choose between two absolute stories: that every grant is benevolent capacity building, or that every recipient has been bought.

The proper entity of analysis is structure. Who defines the funding agenda? Who appoints reviewers? Which applicants hear about the call? How concentrated are repeated awards? What share of a recipient's relevant budget comes from the funder and connected bodies? Does the funder later cite recipients as evidence of regional support? Do recipients enter committees, conferences or elections controlled by the same institutional family? Can a grantee complain, publish an unfavourable finding or decline a publicity request without risking payment?

These questions are especially important for a regional Internet registry. A registry occupies a durable coordination position and collects recurring fees from networks that cannot casually choose another regional allocator. Its technical reputation can open doors for grant recipients. Its conferences and policy structures can give them visibility. The same grants that correct market failures can therefore extend the institution's social reach beyond the narrow service for which members pay.

The right response is separation, not suspicion. Public benefit should be evaluated independently. Institutional participation should remain open on equal terms. Financial relationships should be visible when recipients speak, vote or advise on matters touching the fund. A recipient should never have to perform loyalty, and the institution should never be able to convert a funded project list into a constituency list.

APNIC's grant system is an institutional family, not one account

ISIF Asia predates the current APNIC Foundation. The Foundation's institutional history records that ISIF Asia began in 2008 as a partnership involving APNIC, the Internet Society and Canada's International Development Research Centre. APNIC administered it until 2016. The APNIC Foundation, established in September 2016 after Executive Council discussions that began in 2014, became operational in 2017 and assumed administration of the programme.

This history matters because the funding system cannot be understood from one set of accounts. APNIC is the membership corporation and regional registry. The APNIC Foundation is a separate development organisation with its own Board and audited report. The Asia Pacific Internet Development Trust, created through the transfer and sale arrangements for legacy address space, funds Internet development through the Foundation and other channels. Donors and partners can support particular themes.

Each body can have a legitimate function; together they form an institutional family with money, staff, legal services, programmes and reputations moving between them.

The Foundation's public Board page illustrates the overlap. Its directors include people with long records in regional Internet institutions; one serving director is APNIC's General Counsel and advises both bodies; the Foundation's executive director sits ex officio. Expertise and shared legal support can lower cost and preserve knowledge. They also make it important to identify which body a decision-maker serves at each stage and whose interest the person must advance.

The Foundation's audited 2024 annual report records related-party transactions with APNIC Pty Ltd, identifies Ernst & Young as auditor and states that in January 2025 the company entered a grant arrangement under which it could receive up to USD 10 million during the year for approved projects. The report is evidence of financial scale and organisational connection. It does not show that any grant decision was politically directed.

APNIC members nevertheless need a consolidated relationship map. It should show Board appointment rights, shared staff and services, grants received, grants paid, administration fees, intellectual-property ownership, use of the APNIC name, data sharing, secondments, guarantees and outstanding balances. It should distinguish member-fee money from trust income and external donor funds. A donor restriction does not remove accountability; it changes who set the purpose.

Without that map, each body can tell a narrow story. APNIC can say the Foundation is separate. The Foundation can say the Trust or donor supplied the funds. The Trust can say the Foundation delivers the programme. All may be legally accurate while no member-facing account explains who chose the agenda, appointed the people, bore the cost and received the institutional benefit.

Scale creates a regional field, not just a collection of projects

ISIF Asia's cumulative scale is clear in APNIC's 2025 call for applications. As of January that year, it had supported 166 grants and 43 awards, distributing more than USD 10.1 million across 32 economies. Eligible applicants came from public and private organisations, academia, non-profits and social enterprises across 56 economies in APNIC's service region. The programme grouped work under inclusion, infrastructure and knowledge, with additional environmental and IPv6-deployment grants.

Annual rounds are large enough to influence what organisations propose. In 2023, APNIC announced 24 grants worth a combined USD 2,284,109 across 17 economies. The portfolio included infrastructure, inclusion and knowledge projects, with five IPv6-deployment awards and work on research networks, connectivity, disaster resilience, disability access, digital opportunity and routing. In 2024, 21 projects across 17 economies received USD 1.675 million, including six infrastructure grants, eleven inclusion grants and three knowledge grants, as described in the published announcement.

These numbers demonstrate activity and geographic spread. They do not reveal how important the programme is to the recipient field. Ten million dollars across more than a decade may be modest relative to telecommunications investment, yet decisive for small research groups, operator communities and non-profits. A grant of USD 50,000 can finance a post that otherwise does not exist. The funder can therefore shape which questions are researched, which technologies receive demonstration projects and which organisations acquire regional visibility.

Theme design is the first allocation decision. A call for IPv6 deployment will produce IPv6 proposals; a call for inclusion will invite applicants to describe work through inclusion. Categories can correct neglected needs, but they can also pull applicants toward the institution's agenda. The annual report should state who proposed each theme, what evidence of unmet demand was used, whether members or affected communities were consulted, and which credible themes were declined.

Donor alignment adds another layer. Restricted money can expand the total available without drawing on ordinary membership revenue, but the restriction may move agenda-setting from members to a trust, government, company or philanthropic partner. The annual portfolio should therefore show who supplied each funding envelope, which conditions shaped eligibility, whether the donor took part in selection, what branding or access it received and whether unused money could be redirected. A large restricted award is not automatically more independent than a small member-funded one; it answers to a different principal.

Where several donors support a common programme, the Foundation should explain how it prevents the most generous contributor from defining the regional need.

Administration deserves the same visibility. Grant money can be accompanied by due diligence, contracting, safeguarding, currency conversion, technical advice, travel, communications and evaluation. These services can make a project viable, particularly for a first-time recipient, but they also increase the institution's role in shaping execution. Reporting the full cost per programme and the kind of support supplied would let readers distinguish a simple transfer from an intensive intervention.

It would also reveal whether small awards consume disproportionate administration or whether that extra cost is justified by access for organisations excluded from conventional funding.

Geographic reach needs a denominator. Thirty-two economies with at least one historical grant means that twenty-four of the stated 56 had none by the reported date. That fact does not prove exclusion. Some economies may have stronger domestic funding, fewer applicants, legal barriers or projects unsuited to the programme. The fund should publish eligible applications, award rates and amounts by broad subregion and economy-income band so members can distinguish lack of demand, weak proposals and inaccessible programme design.

The unit of analysis should also be the organisation, not only the project. Two projects with different titles can go to the same institution or connected leadership. Recipient concentration, repeat awards and awards to organisations whose officers serve in APNIC, Foundation, Trust or conference roles should be reported. The purpose is not to disqualify experienced organisations. It is to show whether a regional field is becoming dependent on a recurring circle.

Selection begins before the scoring panel meets

Grant governance is often described as a competition: applications arrive, experts score them, winners are announced. The most influential choices occur earlier. The funder selects themes, eligibility, language, deadline, maximum amount, required legal form, evidence burden and dissemination channels. Those choices determine who reaches the panel.

A small operator group may have an excellent infrastructure proposal and no grant writer. A university with a development office can convert ordinary work into the vocabulary of a call. An organisation already known to Foundation staff may learn which interpretation of a category is likely to succeed. A new applicant may spend days registering only to discover that an approval step takes two working days near the deadline, as the 2025 call warned. None of these differences necessarily reflects favouritism. They shape competition all the same.

The public selection record should therefore start with outreach. Which economies, languages and sectors received direct notice? Were applicant briefings recorded and open? Were questions and answers shared with everyone? Did staff offer proposal advice, and if so under what rule? Assistance can widen access, but private coaching creates unequal information.

Scoring criteria should separate four judgments. Eligibility asks whether the applicant and proposal fit the rules. Technical merit asks whether the method can work. Public value asks whether the expected benefit justifies cost. Portfolio judgment asks whether the award complements other projects, regions and risks. Combining all four in one unexplained score lets a panel reject a strong project for portfolio reasons while implying that its quality was weak.

Reviewer identity and conflict management matter. A small Internet community makes complete distance unrealistic. Experts will know applicants, have served on the same committees or work for connected organisations. The answer is disclosure, recusal and replacement, not a claim that familiarity does not matter. The public report can list reviewers and aggregate recusals without exposing confidential applications.

Decision-makers should record reasons for deviations from rank order. A Board may reasonably change a panel recommendation because of legal, sanctions, security, concentration or donor restrictions. It should state the category of reason and who voted. Hidden executive discretion turns technical scoring into theatre.

Unsuccessful applicants need procedural dignity. They should receive the scored criteria, a concise reason and a route to correct factual or conflict errors, while accepting that funding judgment is not a legal entitlement. Aggregate rejection reasons help the programme improve. A list of winners alone tells the public what the institution celebrates, not how it chose.

Invitation-only money needs its own constitution

The APNIC Foundation's Trust Discretionary Fund is explicit that applications are by invitation. The fund supports initiatives, events, papers, research and online services on behalf of the Asia Pacific Internet Development Trust, while directing organisations interested in open applications toward ISIF Asia.

Invitation can be a legitimate instrument. An urgent technical study may need a known expert. A fragile local institution may be unable to compete in a public call. A small intervention can lose its value during a long application cycle. A donor may authorize a narrow purpose for which only a few organisations are capable. Discretion is not evidence of abuse.

It does change the accountability point. In an open call, scrutiny focuses on selection among applicants. In an invited fund, it must begin with how candidates entered the consideration set. Who identified the need? Who produced the longlist? Which connected organisations were considered? Why was one invited and another not? What price or budget comparison occurred? Was the recipient already funded elsewhere in the institutional family?

An invitation register should publish the recipient, amount, purpose, initiating body, selecting authority, method used to identify the recipient, conflicts and expected output. Sensitive emergency support can be disclosed after the risk passes. Where only one organisation was considered, the record should explain unique capability. Where several were approached, the report can identify the method without exposing losing proposals.

Discretionary support for events and community institutions poses a particular influence risk. A grant can keep a local operator group alive and create a valuable neutral forum. It can also make the group's organisers reluctant to challenge the funder or give the funder recurring stage access. Agreements should state that the recipient retains programme independence, may criticize the funder and is not required to endorse policies. Branding and speaking benefits should be valued and disclosed rather than treated as incidental gratitude.

Invitation-only funding should have an annual ceiling and expiry rule. Repeated support to the same organisation should move to an open or competitively tested arrangement unless continuity has a documented reason. Otherwise an agile exception becomes a parallel patronage system.

The open and invited portfolios should be reported together. Moving a familiar organisation between them can otherwise conceal concentration. Members need to see total connected support, including grants, sponsorships, fellowships, travel, service contracts and in-kind assistance. Each category has a different purpose, but all can contribute to dependence and access.

Dependence is a ratio, a sequence and an exit problem

No single grant amount shows dependence. USD 100,000 may be marginal to a university and existential to a small non-profit. The relevant denominator is the recipient's programme or organisational income, adjusted for connected funding.

Recipients should disclose, confidentially where necessary, the share of relevant annual expenditure financed by APNIC, the Foundation, the Trust and closely connected donors administered through them. Public reporting can group bands: below 10 per cent, 10-25 per cent, 25-50 per cent and above 50 per cent. High dependence is not a disqualification. It triggers continuity and voice protections.

Sequence matters too. A one-year project can be followed by a grant extension, travel support, a conference invitation, a consultancy and another award under a different programme. Each decision may be reasonable. Together they can create an employment-like relationship without the protections, procurement and disclosure of employment. A connected-support register should follow organisations and beneficially related leadership across programmes.

Repeat funding deserves a transition plan. The first grant may test an idea; the second may scale it; the third should normally demonstrate diversified revenue, local ownership, service income, government adoption, another donor or a clear reason continued subsidy is the public-interest model. Some public goods will never become commercially self-supporting. The funder should say so instead of promising sustainability as ritual language.

Exit is the strongest test of recipient independence. Can the organisation continue core work when funding ends? Can it retain data, equipment, code and trained staff? Are users warned about service closure? Can another funder support the project without APNIC approval? Does the agreement permit honest criticism in final reports? A grant that produces a permanent service with no maintenance plan can leave communities worse off when it disappears.

The funder also needs an exit. It should be able to stop a project for non-performance, fraud, safety or changed feasibility through a fair, documented process. It should not terminate because a recipient criticizes institutional policy. Appeals and whistleblowing routes should bypass the programme officer where the complaint concerns that officer.

Dependence can be reduced by paying against verifiable milestones without making payment contingent on flattering conclusions. Research grants should protect publication of negative findings. Infrastructure grants should specify technical and financial outputs. Inclusion projects should protect beneficiaries from abrupt withdrawal. Final payments should depend on delivery and account, not on attendance at the funder's conference.

Later participation must be measured without presuming loyalty

The title "future supporter" describes a governance hypothesis, not a verdict about recipients. To test it, the institution would need to observe what happens after an award and compare it with a credible baseline.

Possible later roles include speaking at APNIC or Foundation events, serving on grant panels, joining advisory groups, holding Special Interest Group office, standing in Executive Council or Number Council elections, nominating candidates, appearing in public consultations, becoming a contractor, or receiving further travel and sponsorship. Public praise and criticism can also be recorded when it relates to institutional decisions. These events show association, not motive.

A simple before-and-after count would be misleading. Grant recipients are selected partly because they already work in the Internet community, so they may have participated before funding. Strong projects can legitimately produce more invitations. People who value APNIC's mission are more likely both to apply and to speak favourably. Selection effects can create an apparent loyalty pattern even if the grant changes no opinion.

A better evaluation would record pre-grant participation, compare recipients with eligible unsuccessful applicants or similar organisations, and examine changes over several years. It should separate roles offered by the institution from actions initiated by the recipient. Being invited to speak is evidence of institutional access; voting for a candidate is an exercise of member rights. The two should not be collapsed.

Privacy limits are essential. The objective is not to publish a political dossier on every grantee. Organisation-level grants and public offices can be matched publicly. Individual views should be aggregated unless the person speaks in an official public role. Ballot secrecy must remain absolute. APNIC or another registry should never use confidential voting, membership or account data to infer recipient loyalty.

Interpretation should be cautious. If recipients later serve at a higher rate, that could mean the grant built useful capability, the institution recruited from a familiar pool, or dependence influenced behaviour. Interviews, appointment records and comparison groups can narrow the possibilities but may not prove causation. The report should state uncertainty rather than choose the explanation most convenient to management or critics.

The institution should also count dissent. Did grantees publish adverse findings, oppose a proposal, decline branding or complain about administration without losing access? Visible disagreement is evidence that the relationship can tolerate independence. It should not become a quota or performance target; manufactured dissent would be as misleading as manufactured praise.

The central boundary is authority. A recipient can support APNIC as an individual or member. APNIC cannot cite the recipient's grant-funded presence as independent regional consent without disclosing the relationship and the basis for any representative claim.

Project completion is not public impact

Grant reports naturally focus on delivery: equipment purchased, code released, people trained, events held, papers written and funds spent. These outputs matter because money should produce what the agreement promised. They are not the final public-value test.

Infrastructure requires operation evidence. Was the exchange point, IPv6 deployment, community network or security service working after twelve and twenty-four months? How many networks or users adopted it? What latency, resilience, route security or cost condition changed? Who pays maintenance? Were procurement and configuration secure? A photograph of installed equipment proves installation, not durable use.

Training requires competence and application. Attendance and completion show exposure. Pre- and post-assessment can show immediate learning. Later sampling can show retained skill and whether entities changed a network, taught others or entered relevant work. The programme should not demand sensitive configurations or personal career data. Aggregated evidence can distinguish a course from a capability outcome.

Research requires methods and intellectual independence. The final report should publish the question, data, limitations, negative findings and reusable outputs where safety permits. Citation and conference presentation are useful, but the harder question is whether operators, policymakers or researchers could examine and use the work. A study that confirms the funder's preferred position should receive no easier path than one that challenges it.

Inclusion projects need a beneficiary denominator and a harm test. How many intended users existed, how many used the service repeatedly, who remained excluded and why? Did the project collect sensitive data, create surveillance risk or expose vulnerable groups? A compelling story about one beneficiary can explain experience; it cannot establish population impact.

Environmental and disaster-resilience grants need physical measures and maintenance periods appropriate to the intervention. A sensor project should show uptime, accuracy, response use and replacement responsibility. A low-power network should account for embodied and operating effects rather than label any digital activity green. The funder need not impose one metric across unlike projects, but each award should define a falsifiable result before money is committed.

Independent evaluation should sample projects, including failures and small awards. Programme staff can monitor delivery; an evaluator should test whether outcome claims follow from evidence. Recipients should have a right to correct facts and respond to judgments, not to suppress an unfavourable evaluation.

Cost belongs in the analysis. Report administration, technical assistance, travel and evaluation alongside grant payments. A smaller grant with intensive support may be good value, but members and donors should see the full intervention cost. The result is a portfolio account that can compare different routes to the same public good rather than celebrating money distributed.

The RIPE NCC decision to stop is a useful warning

An institution rarely provides a clearer exhibit of grant limits than the RIPE NCC did in 2025. Its Community Projects Fund had begun in 2016, distributed about EUR 1.6 million to nearly 50 projects over nine years, paused its 2025 cycle and was then ended permanently by the Executive Board.

The minutes of the 183rd Executive Board meeting are valuable because they record difficulty rather than only success. Management identified financial questions, geographic participation and measurement of success as major challenges. A Board member observed that the range of proposals made value difficult to judge and that measuring success was almost impossible. The Board considered alternatives, including targeted research, underserved-area support and cessation, and unanimously chose to end the fund.

This decision does not prove that the funded projects lacked value. The registry's public notice says existing 2024 commitments would continue and acknowledges recipients and volunteer selectors. Nor does it establish that every RIR grant should stop. It demonstrates that broad mission language such as the good of the Internet cannot solve portfolio comparison, member-benefit and evaluation problems.

There are two lessons for APNIC. First, scale does not cure evaluability. A mature programme with many recipients can accumulate stories faster than comparable evidence. Categories as broad as inclusion, infrastructure and knowledge need distinct outcome methods and a portfolio rule for choosing among them.

Second, stopping is a governance capability. Programmes often survive because recipients, staff, reviewers and conference communities have invested in them. Those interests can be honourable and still resist reallocation. A published expiry, periodic independent review and transition plan make cessation less arbitrary. Existing recipients should be protected from retrospective cancellation when they have performed; future rounds should not become an entitlement.

ARIN's Board minutes offer a smaller procedural exhibit. In 2022 the Grant Committee reviewed 14 applications and recommended three projects. During Board discussion, the chair asked whether conflicts were considered and management said they were; the treasurer also asked why 19 apparent applications became 14 full reviews. The explanation referred to missed contacts, deadlines and communication. The exchange shows the value of recording the denominator and conflicts at the approval stage.

Comparative evidence should not become a league table. APNIC Foundation, RIPE NCC, ARIN and LACNIC programmes differ in funding source, region, legal form, category and scale. The transferable lesson is that a grant portfolio needs a way to learn, narrow and end, not merely a way to announce winners.

Conflicts arise on both sides of the cheque

Grant conflict policies often focus on reviewers who know applicants. The recipient side deserves equal clarity.

A reviewer should disclose employment, consultancy, Board service, family relationship, active collaboration, financial interest, recent supervision and substantial public advocacy for an applicant. The programme should define the period and consequence. Recusal should remove access to scoring and deliberation, not merely the final vote. Because small technical communities have dense relationships, the report should show how replacement expertise was obtained.

Staff conflicts include programme officers who helped design an application, managers whose targets depend on spending the budget and executives who want a flagship project in a preferred economy. Donor representatives may have commercial or geopolitical interests in a theme. Board members may serve organisations connected to recipients. A complete register identifies the role, declared interest, mitigation and decision authority.

Recipient conflicts arise when a grantee later helps allocate the same fund, evaluates competitors or oversees the programme. A cooling-off period should normally prevent a current grantee and key project officer from serving on selection or evaluation bodies for the same programme. Afterward, service can be valuable if the previous award and recusal rules are public.

Institutional governance conflicts require narrower treatment. A grantee should not be barred from ordinary policy participation, voting or standing for office merely because it received funds. Such a ban would let the funder reduce member rights by writing a cheque. Instead, candidates and officeholders should disclose material current grants, and recuse from decisions directly affecting their award, programme renewal or unresolved dispute.

Public communications create another conflict. A recipient may be asked for a quote, case study, panel appearance or ministerial meeting. These can disseminate valuable work. The grant agreement should say that publicity is voluntary except for reasonable factual reporting, that the recipient may approve attributed quotations, and that refusing praise will not affect payment or future eligibility.

Data is a final boundary. The funder may hold budgets, staff details, beneficiary information and project difficulties. Those records must not be used to pressure a recipient in unrelated governance debate. Access should be limited by purpose, with retention and deletion rules. Aggregate evaluation should protect vulnerable beneficiaries and commercially sensitive network data.

A conflict register will reveal many relationships. That is evidence of a connected field, not automatic misconduct. The governance result lies in how consistently interests are disclosed, decisions are reassigned and criticism remains possible.

A grant constitution should protect independent recipients

The most effective reforms can be assembled into a grant constitution applying across the institutional family.

First, publish an authority map. It should identify the legal entity, funding source, Board appointment rights, programme owner, staff support, final decision-maker, appeal body and financial-account location for every fund. Related-party services and administration cost should reconcile to audited reports.

Second, publish the full selection design before applications open: themes, evidence for need, eligibility, scoring, portfolio criteria, reviewer appointment, conflicts, treatment of staff advice, due-diligence grounds, decision authority, complaint route and timetable. Material changes should be announced to all applicants.

Third, maintain one connected-support register. For each recipient organisation it should aggregate grants, sponsorship, fellowships, travel, consultancy and in-kind support from APNIC, the Foundation, the Trust and programmes they administer. Amount, purpose, year, selection route and current status should be visible. Personal beneficiary details should be protected.

Fourth, require an outcome plan proportionate to project type. It should state baseline, intended beneficiaries, denominator, output, later outcome, evidence method, risk, maintenance owner, publication rule and review dates. Failure can be an acceptable result if honestly reported and learned from; unsupported success cannot.

Fifth, separate grant administration from governance claims. Recipient counts, quotes and attendance should not be used as proof of regional consent. Any public statement implying representation should identify the appointing authority and disclose relevant financial support.

Sixth, protect voice. Agreements should guarantee the right to publish methodologically sound adverse findings, criticize institutional policy, decline promotional activity and use ordinary member rights. Payment decisions should be based on agreed delivery and lawful conduct. Complaints about retaliation should reach an independent Board or ethics channel.

Seventh, impose review and expiry. Open programmes should receive periodic independent portfolio evaluation; invited funds should have annual ceilings and a stricter recipient-identity record. Repeated awards should require an exit or continuing-public-good case. Board papers should present stop, narrow, redesign and continue options.

Eighth, report later participation carefully. Aggregate public data can show recipient organisations taking institutional roles, before-and-after patterns, invitations, repeat support and visible dissent. It should never compromise secret ballots or create files on private political opinion.

These controls do not make grants cold or adversarial. They protect the relationship from the expectation of gratitude. A recipient can thank the funder for enabling useful work and oppose it the next day on a fee, election or policy question without either act being treated as evidence of bad faith.

Fund outcomes, not a constituency

Internet-development grants occupy a difficult space between technical coordination, philanthropy and institution building. Their strongest defence is not that the funder has good intentions. It is that a defined public problem was addressed at proportionate cost, through fair selection, with evidence that survives the publicity cycle.

APNIC and its Foundation have helped create a substantial regional grant field. The cumulative record of ISIF Asia, the annual amounts and the diversity of projects show why the programme matters. The Trust adds resources and invitation-based flexibility. The institutional family also creates a larger surface on which money, appointments, events and later participation can interact.

That interaction should not be narrated as scandal without evidence. A grantee who later supports APNIC may have supported it all along, may have been persuaded by its performance or may simply agree on one question. A critic who receives money may remain independent. A project can be valuable even when its leaders enter institutional office. Causal claims require more than a timeline.

Members should still refuse a convenient fiction: that funding has no political effect because no one asked for a vote. Money changes capacity, visibility, access and perceived alternatives. It can make an organisation stronger and freer; it can also make loss of favour more expensive. A governing body that benefits from the resulting network should examine the pattern rather than cite beneficiaries as a self-validating community.

The decisive separation is between outcome and constituency. Project evaluation asks whether infrastructure worked, skills endured, research informed practice and intended users benefited. Governance reporting asks whether selection was fair, support concentrated, recipients became dependent, conflicts were handled and later participation was represented accurately. Neither account should be allowed to substitute for the other.

A successful programme should be willing to fund work that proves a favoured assumption wrong, support an organisation that later criticizes it, disclose a failed project, decline a familiar applicant and end a category that no longer earns its cost. Those are signs that the grant serves a public purpose rather than recruiting agreement.

The Internet community does not need recipients to pretend that money leaves relationships untouched. It needs institutions to design relationships in which gratitude carries no obligation, criticism carries no penalty and a project list cannot be mistaken for a political base. Fund the future, certainly. Do not claim to own its supporters.

Sources