Written by Dr. Rishi Gulati, Associate Professor of International Law at UEA Law School and a Barrier specializing in international law

1 Introduction

Multilateral development banks (MDBs) such as the World Bank, the Asian Development Bank, and the African Development Bank finance large-scale public projects globally and regionally to promote sustainable development and reduce poverty. MDBs require the adherence to environmental and social standards to mitigate harm. However, during project implementation these standards are not always respected and communities may be harmed.  Without timely access to justice for affected individuals, basic rights would be compromised, severely undermining the accountability of MDBs as public international organizations (IOs).

IO accountability has been a significant international legal challenge. Fundamentally, due to IO immunities, national courts are not a viable forum for victims to seek justice against IOs. And international or institutional mechanisms are often unavailable or deficient. To give credit where it is due, MDBs have taken important steps to enhance their accountability. In a landmark development, in 1993, the World Bank Inspection Panel was established and is intended to constitute “an independent complaints mechanism for people and communities who believe that they have been, or are likely to be, adversely affected by a World Bank-funded project.” The Inspection Panel exercises its mandate in the context of the International Bank for Reconstruction and Development (IBRD) and the International Development Association (IDA), both World Bank institutions.

In 1999, the International Finance Corporation (IFC) and the Multilateral Investment Guarantee Agency (MIGA), also World Bank organizations, created the Compliance Advisor Ombudsman (CAO) as its accountability mechanism. Most other MDBs as well as bilateral development institutions followed suit and there are currently more than 20 such independent accountability mechanisms (IAMs) in place (August Reinisch, Third report on the settlement of disputes to which international organizations are parties, paras. 48-58). However, as is demonstrated by the example of the Cambodia micro-finance case examined by the CAO, there remain serious challenges in the effectiveness of such bodies.

2 The Cambodia micro-finance case and the CAO’s findings

In February 2022, two Cambodian human rights organizations (LICADHO and Equitable Cambodia) filed a complaint with the CAO, on behalf of a group of individual microfinance borrowers who took loans with one or more IFC client/sub-client. The complaint concerned IFC’s investments in six Cambodian microfinance lenders and alleged that their lending and debt-collection practices had caused over-indebtedness, forced sale of Indigenous lands used as collateral, food insecurity, and intimidation of borrowers.

IFC Management purported that impact on micro-finance borrowers and financial consumer protection fell outside its Environmental and Social (E&S) Sustainability Framework, and therefore outside the CAO’s mandate. In June 2023, following its preliminary compliance appraisal in which the CAO decided to initiate a compliance review into the case, IFC submitted a request asking the IFC Board of Directors to overturn the CAO’s decision to investigate, arguing that CAO’s interpretation that all E&S issues fall within IFC’s Sustainability Framework is “expansionary and unwarranted.” However, IFC Management withdrew the request a month later, and the investigation went ahead without a formal challenge at that stage.

In October 2025, the CAO published its investigation report. CAO found no basis for excluding sub-borrowers from the Sustainability Framework, which applies to a project’s full range of E&S risks “unless specifically excluded.” The CAO determined that the social risks of microfinance lending fall within the scope of IFC’s E&S requirements. It found that IFC had failed to meet its obligations under Performance Standard 1 in lacking adequate E&S oversight of client lending practices and impacts on vulnerable groups, and under Performance Standard 7 by allowing the use of Indigenous communal land as collateral without community consent and in lacking culturally appropriate products and communication by IFC’s clients. CAO further found that IFC had not ensured that vulnerable people are not disproportionately affected by project impacts.

3 The actions of the IFC Board

As background, it is worth noting that IFC activities are guided by a Board of Governors and a Board of 25 Directors (the “Board”), with the former  delegating most of their powers to the latter. The CAO’s Policy, which governs the mandate and procedures of the CAO and the relationship to the Board, was adopted by the Boards of IFC and MIGA, establishing CAO as an independent recourse and accountability mechanism.

According to the CAO Policy, once CAO’s investigation report has been finalized following its consideration of comments from both IFC/MIGA Management and the Complainants, the final investigation report is “submitted to Management and circulated to the Board for information” (emphasis added). The Policy then provides that, within 50 business days of receiving CAO’s compliance investigation report and findings, Management will be required to submit a report to the Board, stating proposed actions in response to the CAO’s findings. The resulting Management Action Plan is then subject to Board approval.

Nevertheless, in the present instance, the IFC Board’s response to CAO’s investigation report departed from the procedure as set out in the CAO Policy. On 24 June 2026, the IFC Board issued a public statement, noting that the IFC Board approved Management’s report including a “Special Management Action Plan”, and that in approving the plan, the Board “recognized that there has been no policy noncompliance under IFC’s Policy on Environmental and Social Sustainability.”On 6 July 2026, the IFC Board further instructed CAO to suspend processing of ongoing CAO cases relating to financial consumer protection in microfinance and to not accept any new such cases, pending discussion on the policy issues. This decision affected 10 ongoing CAO cases in various stages of processing, from initial assessment, compliance appraisal and dispute resolution. These Board decisions contradict the CAO policy which lays out that it is in the sole authority of the CAO to admit complaints and no interruption of ongoing cases through a Board intervention is provided for in the policy.

4 Accountability undermined – independence challenged

The Board’s actions have caused strong reactions. The CAO’s Director General resigned from her role. Civil society organizations and experts responded with an open letter signed by 91 organizations and eight renowned experts in the field of IAMs, warning that the decision “sets a dangerous precedent for accountability at the World Bank Group.”

The Cambodia case brings out a structural tension in the design of IAMs and their relationship with the Board of the MDB. If IAMs are to be effective, they must be independent from both Management and the Board. A critical aspect of independence requires that the IAMs determinations are respected and not undermined by the executive organ of an IO. IAMs are designed to provide impartial, professional and evidence-based compliance oversight, insulated from political considerations, whereas the Board performs a dual function: overseeing Management while also serving as representative bodies of the member states. The IAMs provide the Board with credible compliance findings, to serve their oversight over Management. In fact, an effective IAM improves project outcomes.

In the present circumstances, it is immediately apparent that first, the Board departed from the CAO Policy, which it must not do. Such non-compliance manifestly violates the internal law of the organization.  Second, it is suggested that the Board’s actions undermine the very reasons for which the CAO was created, as well as its functions and mandate. Amongst other things, the CAO was created to facilitate the resolution of complaints related to IFC/MIGA projects, carry out investigations, facilitate “access to remedy for Project-affected people” in “a manner that is fair, objective, and constructive”, and foster public accountability (see especially, CAO Policy, paras. 4, 5 and 7). The Board does not have the authority to arbitrarily ignore or substitute the CAO’s findings; on the contrary, it may strengthen implementation of remedial measures (CAO Policy, paras. 128, 137 and 144). The Board’s response in the Cambodia micro-finance cases does not comply with the letter and spirit of IFC’s own internal legal framework.

Ensuring an effective IAM is not only important for project effectiveness, necessary to ensure compliance with institutional E&S commitments, but also to reduce external legal risk for an MDB. As the Jam et al. v. IFC litigation has demonstrated, MDBs’ immunities are starting to be restricted domestically.

Moreover, recent jurisprudence from the United Nations Human Rights Committee in MLD v Philippines, Communication No 3581/2019 (18 September 2024) has globalised the European Court of Human Rights’ Waite And Kenedy jurisprudence by saying that IOs must provide “reasonable alternative means” of resolving disputes to private persons they have allegedly harm. This jurisprudence is in the labour context, but it can readily be applied to claims by other third parties as well. While an IO has “discretion as to the reasonable alternative means of dispute resolution”, such means must be  independent, impartial, and not amount to “arbitrariness or denial of justice” (MLD v Philippines, para. 9.7). If such “reasonable alternative means” of dispute resolution are not provided, then there is a real risk that IO immunities will be breached.

In fact, in its project on the “settlement of disputes to which international organizations are parties”, the International Law Commission is placing a central emphasis on independence and impartiality as constituting key features of procedural justice and the rule of law (see especially, August Reinisch, para. 223). I have argued elsewhere that access to justice requires that the dispute resolution forum created by IOs is able to deliver justice independently, impartially and fairly, and that, amongst other things, this requires that the mechanism be able to issue binding decisions. A frequent criticism of IAMs is that their decisions are not binding, unlike other review mechanisms which issue binding decisions towards the same institutions, such as administrative tribunal for personnel-related claims, or anti-corruption investigations and sanctions.

5 Concluding comments

Ultimately, an IAM’s legitimacy depends on its independence from Management, as well as on the credibility and predictability of its procedures. IAMs are designed to serve as impartial, evidence- and fact-driven arbiters between Management and external affected parties. MDB Boards are not equipped to perform this function and their intervention in an IAM’s findings or procedures may cause mistrust and undermine the IAM’s credibility, and by extension the credibility of the institution’s E&S commitments.

The lack of acceptance of the independent findings of an IAM such as the CAO does not only undermine its effectiveness and contradict internal regulatory frameworks, but it also raises legal risks that immunities could be breached. This would be a highly undesirable outcome for IOs as well as their member states. Accordingly, regardless of their origin, if they are to be considered a “reasonable alternative means” to resolve disputes between the institution and external parties, IAMs must be able to deliver justice independently and impartially, with an ability to make final compliance determinations which are not open for reconsideration on the merits at the Board level.

The World Bank Group is currently engaged in a restructuring process integrating its three independent accountability mechanisms comprised of the Inspection Panel and the Dispute Resolution Service, as well as the CAO. The aim is to create a single IAM and avoid fragmentation. On its face, creating a streamlined mechanism may seem to be a positive development. However, given the geopolitical circumstances, and the weakening of the rule of law internationally, it is crucial to remain vigilant that this restructuring process does not weaken the existing legal infrastructure in place. As I have argued in this post, MDBs should create IAMs whose decisions are final and binding, where such mechanisms are of a judicial character. Realistically, at the very least, it would be a great disservice to the cause of MDB accountability if existing IAMs are undermined by diluting or ignoring their policies and procedures.