A Symphony of Institutional Failures: The Venezi Affair

July 10, 2026

Germana Barba

Imagine ninety professional musicians learning from a press release that they have a new boss. That is what happened at Venice's Teatro La Fenice in September 2025, when superintendent Nicola Colabianchi announced the appointment of conductor Beatrice Venezi as Music Director without any prior consultation with the orchestra. Within three days, all ninety members had signed an open letter demanding the immediate revocation of the nomination. And they were not alone: all 300 workers of the Fondazione unanimously voted to support them, denouncing "procedures and timing that trampled every principle of consultation and transparency." By April 2026, the Fondazione had formally terminated its future collaboration with Venezi, citing a statement made by Venezi to an Argentinian media outlet that orchestra positions were "handed down from generation to generation."

Most commentators focused on the political dimension: Venezi is known to be a personal friend of Italy’s Prime Minister Giorgia Meloni and close to her political party. But the affair raises a problem that goes deeper than political patronage. In a democratic system, institutions exist to serve people who have a legitimate stake in how power is exercised; people who are entitled to ask for reasons, and to receive them. When an officeholder acts without giving those reasons, they are not merely cutting corners: they are treating those affected as if their claims did not matter. That is a form of injustice in the relationship between the institution and those it serves, regardless of whether any rule has been formally broken.

Although general managers are legally entitled to appoint a music director unilaterally, it is standard professional practice to consult the musicians first, and for good reason. The musicians are not peripheral stakeholders; they are the people who execute the institution’s artistic mission.  The content of the orchestra's protest makes this even clearer. Their open letter was explicit: "Our opposition derives exclusively from the professional profile of the appointed music director." The musicians noted that Venezi had never conducted an opera or a public symphonic concert at La Fenice, and that her curriculum was "not remotely comparable" to that of conductors who had previously held the role. Furthermore, subscription cancellations multiplied in the days following the announcement, suggesting that audiences too felt that something had been taken from them; that they, as the public that a publicly funded theatre exists to serve, also had a claim to be considered.

The reversal of April 2026 did not fix this. The Fondazione ended the collaboration without acknowledging that the original process had been flawed. That matters. Accountability broke a second time when no explanation was given to those who were bypassed. Without that, the reversal is simply another unilateral act. The Venezi affair is therefore not a closed case. It is a reminder that the way institutional decisions are made (who is consulted, who is owed reasons, who gets to ask why) is as crucial as the decisions themselves.

How to Restore Trust After a Falsified Election? Lessons from Vernier, 2025

July 03, 2026

Matthieu Debief

In March 2025, the commune of Vernier, in the canton of Geneva, voted to elect its municipal council. Three months later, the result was thrown out. Handwriting analysis showed that 278 postal ballots had been filled in by the same nine people. In June, Geneva's Constitutional Chamber annulled the vote, saying the outcome could no longer reflect "the faithful and assured will" of the electorate. A fresh election in November raised renewed suspicions of fraud, but this time the commune let the result stand.

Electoral fraud might look like a one-off crime that a new election can fix, but it wounds something no democracy can do without: trust. And an election is built on trust more than almost any other public act. In the canton of Geneva, more than 75% of votes are cast by post where there is no register of signatures and no systematic identity check, so the system takes the voter's good faith on trust. That is exactly what the Vernier fraud exploited.

The concept of trust has, at least, two dimensions. The first is external: the confidence citizens place in the institution from the outside. In an ethic of public office, power is not owned, it is entrusted. An officeholder receives it to carry out a mandate and must account for how it is used. An election is the moment that delegation is made. Citizens hand over power on the understanding that it was won fairly, and electoral fraud breaks that understanding. In Vernier, it happened on already thin ground: turnout in communal elections barely tops 30%, leaving the mandate resting on a narrow slice of the electorate.

The second dimension is internal: the trust elected officials must place in one another. To govern as a body, each one must assume that their colleagues hold their seats legitimately and use the powers of office for the job, not for themselves. In Vernier, the fraud poisoned the makeup of the council itself. An assembly that cannot vouch for how its members got there cannot really act as one.

Each dimension of trust needs a different fix. For the external one, the answer is not necessarily to get rid of postal voting but to make the process trustworthy. Cancelling the tainted result and voting again was a good first step: it shows the system can correct itself. But a second vote only rebuilds outside trust if it takes away what the fraud was meant to gain, and that is far from guaranteed. In fact, voters tend to punish even proven wrongdoing only mildly, and the Canton of Geneva knows that too well. So, a second vote cannot do the work on its own.

It also needs institutional safeguards around it: impartial administration, traceable ballots, and a count that is open and potentially independently checked. Without those, a second election just launders the first and gives citizens no real reason to trust the outcome.

The fix for internal trust is less direct. After the second election in Vernier, the newly elected members had to work together while doubts still hung over the election. What could restore office trust among them is the idea of office accountability. To trust a colleague is not just betting they will follow the rules. It is relying on their commitment to act for the reasons the office exists in the first place. Accountability is what brings this commitment to light: when each member must justify how they use their power by pointing to their mandate rather than themselves, the pro-office attitude becomes testable rather than merely assumed.

In practice, this can take the form of minimum rules regarding the reporting of conflicts of interest and disclosure, but above all, it also requires elements that rules alone cannot provide: a code of conduct the members write and revise themselves rather than receive from above, training that gets them to see their roles as bound together, and habits of mutual answerability through which colleagues check one another early rather than look away. And once that internal discipline is seen to hold, it can feed back into external trust too: citizens give their confidence to a council whose members clearly hold one another to the job.

The damage from the Vernier fraud was not merely legal but a matter of trust, and trust is what institutions, public ones included, are built on. So, the real cost of for Vernier is not 278 ballots. It is the doubt now hanging over every honest ballot, and every honest mandate, in a canton that can least afford to lose more.

When Corruption Became a Governance Issue

June 26, 2026

Maria Paola Ferretti

In the language of the United Nations, for a long time, corruption was treated mainly as a legal or economic problem: bribery, embezzlement, and the abuse of public office. It was something to be prosecuted, punished, and contained. Between 2000 and 2020, however, this language began to shift. Corruption was increasingly framed in broader political terms, and “good governance” emerged as a central organizing concept in global anti-corruption discourse.

Recent scholarship on the “Rise and Fall of Liberal Aspirations” in international politics highlights how “good governance” became a flexible but powerful umbrella term. Up to the 2000s it was still closely linked to democratic quality: free elections, accountable institutions, media freedom, and an active civil society. Over time, however, its meaning broadened and became more ambiguous, allowing it to be adopted and reinterpreted across diverse political systems, including authoritarian and hybrid regimes.

A key development in this evolution is that corruption came to be positioned as the direct opposite of good governance This shift reflects wider global trends in the early twenty-first century. As authoritarian regimes gained greater influence in international forums, consensus around democracy as a universal benchmark weakened. The language of media freedom, political pluralism, and civic participation became more contested. Yet anti-corruption remained unusually resilient as a shared concern. Even governments with very different political systems could agree that corruption threatens good governance.

In this context, corruption increasingly came to define not just individual wrongdoing but institutional failure itself. Within UN discourse, it began to function as a diagnostic category: if corruption is present, governance is assumed to be weak or dysfunctional. This helps explain why anti-corruption became such a central theme in global policy discussions, even as broader liberal-democratic language lost some of its prominence.

Public debates about corruption today are rarely only about financial misconduct, but primarily about good government. They are fundamentally about trust in institutions, transparency in decision-making, and accountability in the exercise of power. Citizens are asking whether public institutions genuinely serve the collective interest or whether they are captured by private networks and informal influence.

International organizations, development agencies, and NGOs have reinforced this perspective. Transparency reforms, independent oversight bodies, anti-corruption agencies, and accountability mechanisms have been promoted as essential components of sustainable development. The adoption of the UN Convention against Corruption and, later, the Sustainable Development Goals further strengthened the link between anticorruption and good governance.

At the same time, this raises an important question: can the anti-corruption agenda remain effective and normatively relevant if it becomes detached from its democratic foundations? In many UN documents and policy speeches, governance principles are presented as mutually reinforcing. Democracy is said to enable transparency; transparency strengthens accountability; accountability builds trust; and trust stabilizes institutions. Corruption, in turn, is framed as the force that disrupts this entire chain.

So, ultimately this governance-centered framing can sustain its normative only through a renewed connection to democratic language.

 

Figure: Steffek et al. (2025) Mentions of ‘good governance’ (f per million words) and collocations over time, ranked by significance. It is possible to observe the rise in prominence of (anti)corruption starting from 2001.

Can Anti-corruption Become Corrupt?

Anti-corruption and Selective Accountability in Turkey

June 19, 2026

Serkan Seker, Geneva RCCS

Anti-corruption is an exercise of public authority aimed at addressing one major failure of institutional action. As such, it must itself remain subject to accountability, procedural fairness, and public justification.

Turkey’s growing notoriety for corruption, reported in international indices, has intensified as the country has become increasingly authoritarian.  At the same time, particularly since the arrest of İstanbul Mayor Ekrem İmamoğlu in March 2025, the country has been increasingly witnessing “anti-corruption operations” against local administrations run by the main opposition, Republican People’s Party, the CHP. İmamoğlu and his party, who defeated President R. Tayyip Erdoğan’s candidates in the last two Istanbul elections, are seen as the most credible actors against the Turkish president.

However, the point I want to discuss here is not the widespread corruption of institutional action, entrenched in local and national institutions, regardless of officeholders’ political affiliations, across Turkey. And the issue is not whether the allegations should be investigated. Of course they should. The critical point is whether anticorruption is impartially structured and implemented.

What I want to emphasize is that anti-corruption itself must be governed by a public accountability ethos: It should draw on public ethics, honouring the institutional raison d’être and well-functioning. It is fundamental that the authority of accountability enforcement and anti-corruption cannot be exercised selectively by turning into a political instrument for punishing opponents and insulating allies. Therefore, the ethics of anticorruption excludes politically contingent application of rules and norms. Otherwise, anti-corruption itself may become corrupt and corrode institutional mechanisms.

Pro-government media frame the operations as “the anti-corruption campaign of the century” against a supposedly decayed opposition. They publish confidential case materials, including highly personal details and police records, immediately following the operations. Such disclosures appear possible only through collaboration with judicial and administrative officeholders who leak information, seemingly to shape public perceptions before judicial processes unfold.

For circles close to the opposition, the operations are yet another effective instrumentalization of public authority within a broader anti-democratic project, aiming at eliminating the strongest rival in the political struggle. The operations, they argue, aim to recapture the local administrations which the government has lost in the last elections: they are not for preventing corruption and maintaining accountable public institutions. In addition, it is alleged that much broader corruption allegedly committed by pro-government actors have not been investigated.

Anti-corruption is not normatively innocent only because it is semantically associated with accountability. It is not normatively self-justifying. Anti-corruption is an exercise of public authority as well. From the perspective of the ethics of office accountability, officeholders entrusted with the mandate of anti-corruption must remain mutually and publicly accountable.

Anti-corruption policies, most importantly, must be accountable themselves. The ethics of anti-corruption requires a “pro-office” mindset, upholding and maintaining institutional integrity. Such an ethics can serve as the opposite of corrupting institutional action.

Judicial processes, in this vein, must operate according to universal norms: the presumption of innocence and right of privacy must be respected. Judicial power must be fairly exercised towards all actors and parties. It cannot be exercised in a selective manner, depending on officeholders’ political affiliation.

When anti-corruption abandons impartiality, publicity, procedural fairness, justificatory transparency, and non-instrumental use of coercive power, it ceases to operate as a remedy against the corruption of institutional action. Instead, it produces accountability deficits and becomes yet another form of institutional dysfunction.

June 12, 2026

Marta Dainesi, Geneva RCCS

In France, each political party has the right, once a year, to create a parliamentary commission of inquiry on a topic of its choice. Most of them attract little public and media attention. The Commission of inquiry into the neutrality, functioning and financing of public broadcasting was an exception. Its activity and final report generated intense debates and the organisation AC!! Anti-Corruption filed a formal complaint against the Commission’s rapporteur for illegal taking of interests and influence peddling. Why did this Commission become so controversial?

At the centre of the affair stand two figures: the rapporteur, the politician Charles Alloncle, and the businessman Vincent Bolloré, owner of France’s largest private media empire. During the 2024 elections, Bolloré played a key role in the creation and alignment of Alloncle’s party (Union des Droites pour la République) and in his campaign. In the Commission’s proceedings, Bolloré was heard as a major actor in public-private relations in the audiovisual sector, with Alloncle himself leading the session. Alloncle was criticized for asking superficial questions and thereby giving Bolloré a platform to present his views. Bolloré’s influence extended beyond this event: his media outlets promoted Alloncle as a courageous reformer challenging public dysfunctions. Le Monde also revealed that a company controlled by Bolloré had sent questions for the hearings, exceeding officially declared lobbying activities.

Alloncle’s methods also raised concerns. During hearings with key broadcasting actors, he focused on allegations of conflicts of interest rather than devoting attention to the structural challenges of media funding and governance. Concerns about Alloncle’s intentions were reinforced by his membership in a political alliance openly advocating the privatisation of public broadcasting.

This context made the final vote problematic. Rejecting it would have eliminated the data collected by the Commission, depriving the public of valuable accountability resources, while approving it risked legitimizing a report accused of being inaccurate and biased.

This affair raises ethical and legal questions, yet what I find most interesting are not the dysfunctions themselves, but rather the reactions they provoked. Members of the Commission, politicians, researchers, and journalists used interviews on the case not simply to denounce misconduct, but to reflect on the mission of public broadcasting. At a time when political life and public opinion are increasingly shaped by algorithms and external interference, they described public media as a precious democratic reference point. They raised concerns that the proposed reforms, including budget cuts and a market-oriented approach, undermined this mission by reducing pluralism. They also debated how this role could be implemented, which values should guide it (such as neutrality, impartiality, transparency, and pluralism), and how these principles should be formalized in its governance. Despite deep disagreements, they demonstrated answerability by justifying their positions and choices in terms of their understanding of the public mandate.

Journalists also discussed problems within newsrooms. They described a harmful climate, in which pre-emptive editorial caution, aimed at avoiding far-right criticism, risks impoverishing public information and weakening the deeper meaning of pluralism.

Attacks, disagreements, and dilemmas thus helped reveal underlying problems. They pushed individuals to reflect collectively on errors, acknowledge difficulties, and engage critically with their practices and the role of their institution. In this sense, their response went beyond self-justification, opening the way to a reflection on more effective and accountable institutions and stronger safeguards. A dysfunction thus highlighted a democratic and anti-corruption ethic embedded in a network of public actors, one that citizens, to whom public broadcasting ultimately belongs, should recognise and protect. The case of the French public broadcasting Commission reveals a fragile democratic ecosystem. But one that is still alive and functioning.

The Anti-Weaponization Fund: The Undoing of Accountability?

June 5, 2026

Francesco Chiesa, Geneva RCCS

On 18 May 2026, the U.S. Department of Justice announced the creation of the Anti-Weaponization Fund as part of a settlement agreement between President Donald Trump and the Internal Revenue Service. Under the agreement, Trump withdrew a lawsuit concerning the disclosure of his tax returns in exchange for the establishment of a $1.776 billion compensation fund for individuals claiming to have been targeted by politically motivated government action.

The proposal immediately generated controversy. Critics noted that the fund could compensate convicted or prosecuted individuals involved in the January 6 2021 Capitol riot and questioned the concentration of authority over the scheme in officials appointed by the President. Two Capitol Police officers who defended Congress on January 6 subsequently filed a lawsuit describing the arrangement as “the most brazen act of presidential corruption of this century.” The fund has now been put on hold following a judge’s order.

The legal questions surrounding the fund will be resolved by the courts. Yet the case raises a broader and longer-lasting question: can public authority be used to weaken the accountability mechanisms that public authority itself is supposed to uphold?

At the Geneva Research Center for Corruption Studies, corruption is understood as a deficit in the accountability chain linking officeholders to their mandates.  On this view, corruption is not only bribery or personal enrichment. It is the use of the power of office in pursuit of an agenda that cannot be publicly defended as coherent with the legitimate scope of the office’s mandate. From this perspective, corruption is not merely a violation of a rule. It is a distortion of the office itself: holding a public position while turning its purposes inside out.

The controversy surrounding the Anti-Weaponization Fund illustrates why this broader understanding matters. Officially, the taxpayer-funded program is presented as a mechanism to compensate individuals who claim to have been victims of government “weaponization.” Critics, however, argue that its design creates opportunities to reward political allies and individuals closely associated with the president’s political movement, including participants in the January 6 attack. The central issue concerns the institutional architecture through which decisions are made. The problem is not simply that the fund may compensate the wrong people. It is that the design makes it hard to identify who can be held to account for those decisions, and to whom reasons must be given.

The difficulty begins with the settlement itself. A sitting president dropped a personal lawsuit over his own tax returns, and in exchange the government committed $1.776 billion in public money. Whatever the merits of the underlying dispute, the structure is unusual: a private legal interest resolved through a trade that binds public funds.

It deepens with the fund’s governance. The fund would be overseen by commissioners appointed by the acting attorney general, himself a presidential appointee and former lawyer to the president, while the president would retain the power to remove commissioners. Congress plays no direct role in authorizing the scheme. Whether or not the arrangement is legal, it raises a basic question: who can effectively scrutinize the exercise of this power? Moreover, accountability requires more than the possibility of review after the fact. It requires a structure in which officeholders can meaningfully be called upon to justify their decisions to others occupying distinct institutional roles. When authority becomes concentrated within a self-referential chain of appointments, the space for reciprocal justification risks narrowing. Corruption can arise when public power is exercised through arrangements that weaken accountability relationships themselves. This is why the officers’ accusation of corruption is noteworthy. Their claim is not simply that public money may be misspent. It is that presidential power itself is allegedly being exercised outside the standards of accountability that define the office.

Whether the Anti-Weaponization Fund survives judicial scrutiny is therefore only part of the story. The deeper issue is institutional. Accountability begins when public officials govern through structures in which they can be questioned and must answer. Corruption may emerge not only when officeholders stop giving answers, but when institutions are designed so that no one can meaningfully ask the questions. In that sense, the controversy surrounding the Anti-Weaponization Fund is not only about compensation. It is about whether public power can be used to erode the accountability on which public office itself depends.

Patrizia Pedrini, Geneva RCCS

This is not a post about corruption. Nor is it a post about criminal guilt. The Garlasco case—one of Italy’s most controversial murder investigations—is also a story about records, memory, and institutional traceability. What interests me here, however, is only one aspect among many: a form of accountability failure that emerges when the path through which decisions were made becomes difficult to reconstruct. In 2007, a young woman was killed in her family home in the town of Garlasco. Her boyfriend was convicted in 2015, largely on the basis of circumstantial evidence, after years of conflicting judgments and appeals. Now, nearly two decades later, a new investigation started concerning an individual whose position had previously been archived following a rapid supplementary inquiry. This evolution has reopened public debate and may eventually lead to a request for revision of the original conviction. The new investigation has only recently been closed, events remain in flux, and the presumption of innocence fully applies. What interests me here, however, is not guilt but a problem of missing accounts.

In some theories of accountability, institutional failure occurs when officeholders provide an account of their conduct that cannot be vindicated by the terms of their mandate. The justification exists, but fails normatively. The Garlasco affair points toward a more elementary dysfunction: situations in which the account itself is partial, opaque, fragmented—or simply absent. Again and again, across different phases of the investigations, one encounters missing verbalizations, undocumented passages, investigative avenues that appear difficult to reconstruct retrospectively, unclear procedural transitions, and, perhaps most strikingly, repeated “I do not remember” responses from institutional actors asked to reconstruct investigative choices made many years earlier. Some of this is entirely understandable. Time passes, memories fade, personnel changes. But accountability depends precisely on the capacity of institutions to leave behind intelligible traces of reasoning that do not rely exclusively on personal recollection. And that is where the difficulty emerges. In several moments of the affair, the problem is not merely that the institutional account is unpersuasive or controversial. It is that no stable account seems fully available at all. Certain decisions cannot be clearly reconstructed because they were insufficiently verbalized at the time, poorly documented, or never articulated in a way that now permits coherent retrospective explanation.

When this happens, interpretive disorder expands rapidly. The public sphere fills with competing reconstructions, media amplification, forensic disputes, and mutual accusations because the institutional narrative itself appears fractured. Citizens are left not simply doubting conclusions, but struggling to understand how those conclusions were reached in the first place. Importantly, this point does not require allegations of corruption or bad faith. Lack of accountability can emerge well before such thresholds are crossed. Institutions also fail when they cannot provide clear and reconstructible accounts of their own conduct.

The Garlasco case is therefore not only a criminal case. It is also a reminder that accountability depends not merely on acting properly, but on preserving the conditions under which institutional action remains intelligible over time—especially when that action will later need to be explained, scrutinized, and trusted by others.

Legitimizing Dirty Money? The Dangers of Philanthropy Financing Public Institutions

Beatrice Bella, Geneva RCCS

On May 4^th, 2026, the annual Met Gala was held at the Metropolitan Museum of Art (Met) in New York City to raise funds for the Costume Institute. The Costume Institute is part of the Met, though it does not benefit from the Met’s standard endowment. Therefore, it must seek its own funds to realize its raison d’être, which is the valorization and preservation of fashion history by collecting, preserving, studying, and exhibiting clothes and costumes.

This year, the gala drew controversy for its main sponsor: Jeff Bezos, owner of Amazon, and wife Lauren Sànchez Bezos donated an estimated 10 million dollars, raising questions over the role philanthropy should play in financing institutions, particularly when the funds are the result of activities directly in contrast with an institution’s raison d’être. This is arguably the case with the Bezos-Sanchez donation, considering that Amazon has decimated small fashion businesses and mass-marketed cheap labor, fostering a raison d’être of market-driven fashion and profit-centered organization of labor that runs contrary to the values and normative commitments the Costume Institute should uphold through its institutional action.

This controversy could be silenced by arguing that charitable donations, whatever their provenance, are always welcome when they are being used to doing good. The Costume Institute could not function without donations; exhibitions would shrink, conservation works stall, and the public goods the institution exists to preserve would be diminished. A donation that keeps the doors open does real good, regardless of who signs the cheque. If the museum turns the money away, the loss is borne by the public the institution serves, while the donor simply spends it elsewhere. Refusing the gift is a costly gesture.

However, such a consequentialist reasoning would be quite quick. It could also be argued that accepting this kind of philanthropy from companies whose activities run counter to an institution’s raison d’être risks jeopardizing the very nature of the good that the institution is meant to further and only whitewashes the reputation of big donors. It might be worth asking whether Bezos’ donation helped the institutional action of the Costume Institute or ultimately ended up undermining it by pulling focus away. This is not an easy question to settle: the good a donation does and the values it might thus compromise cannot be easily compared.

But a deeper question should be considered too: can an institution still account for what it does once a private donor has helped set its agenda? Philanthropy of this scope is problematic for the unaccountable use it makes of private funds for public goods. The person donating the money has complete control over how much to give, when to give it, and often what is done with it, effectively setting the agenda and orienting institutional action.

Posing that question interrogates the fundamental issue of whether a museum should need private funds in the first place. In the US, only 24% of museum funding is public. These institutions depend on private funds to survive, often staging elaborate and costly fundraisers to keep functioning. The cost of the Met Gala is around 6 million dollars. If, instead of a tax-deductible donation, the money had come from taxes, there would have been no need for a costly fundraiser, and the allocation of the money would have been decided by public democratic standards, subject to normal accountability constraints.

Museums are institutions which provide public goods and services. The fact that a museum does not receive public funds despite offering public goods, but is instead forced to rely on private donations, leaves the door open for “dirty money” and, what is even more ethically worrisome, might legitimize it. Once such a donor helps setting the institution’s agenda, the institution can no longer give a coherent account of its own action by reference to its raison d’être alone. From this vantage point, the accountability conundrum for the Costume Institute is how it can serve its raison d’être of valorizing and preserving the history of fashion with funds generated by activities that arguably undermine it. That is where accountability breaks down—not because the money is morally tainted, but because the donor’s agenda and the institution’s raison d’être are pried apart.

The point, then, is not to look for a different, less controversial billionaire to sustain institutional action. It is to ask what it would take for a museum to be able to account for its own action without depending on whoever happens to be willing to pay for its activities. Reducing that dependence—through public funding, or through endowment structures insulated from any single donor’s agenda—would be challenging, and it is fair to ask who sustains the costs in the meantime. But the difficulty is practical, not principled: as long as institutions that exist to provide public goods are left to stake their survival on private wealth, they will keep being at risk of legitimizing money they cannot fully account for. The Met Gala is not the disease; it is the symptom of a funding model that quietly outsources accountability from institutional action. Naming that clearly is the first step toward changing it.

Dirty Money in Clean Hands

Michele Bocchiola, Geneva RCCS

There is a myth about Swiss banks: impeccable discretion, mountain air, and a vaguely Alpine sense of probity that keeps the dirty money out. But the reality is a little more complicated, as Transparency International noted in 2024: of the 43 corruption cases linked to Africa they were able to follow, 10 led straight to Swiss bank accounts. That is not a catastrophic failure rate, unless, of course, you are one of the countries whose public funds disappeared. So what is going wrong?

A useful way to think about it, I think, comes from thinking about corruption as an unaccountable use of the power of office—when someone in a position of institutional authority acts in ways that cannot be justified with reference to the mandate of that office. On this view, corruption is not only bribe. Clearly, the banker who approves a wealthy client’s account without properly verifying the origin of the funds is not taking a bribe. But they may be exercising discretion in a way that they could never fully account for to a regulator, a judge, or the public. That is precisely where corruption is.

The Swiss Banking Association’s own website seems quite clear: banks can reject clients if they have doubts about identity or the origin of funds. They can ask questions. The problem is that ‘can’ is doing a lot of heavy lifting in that sentence.

FINMA, the Swiss financial regulator, found in its 2024 annual report that a significant number of banks it inspected had anti-money laundering systems that were, to put it mildly, not exactly watertight. The rules exist. The due diligence frameworks exist. And yet the money still finds its way in.

This is the accountability gap. This I not to say that nobody is formally responsible. Indeed, every account is associated to a relationship manager, every institution has a compliance office. The problem is that responsibility is distributed so thinly across forms, procedures, and automated checks that no single person has to own the decision and stand behind it. When things go wrong, the answer is almost always a variation of: ‘the process did not flag it.’ And this is not accountability. This is ‘a whiter shade of pale’ of accountability.

How to address this issue? One natural thought could be to have more rules, and perhaps more severe punishment. But Switzerland already has the rules (so many rules!). What it needs is ‘office accountability’: we need officeholders who can give reasons—bankers and compliance officers who treat due diligence not as a checkbox exercise but as a genuine judgment call, and who would be prepared to show how the rationale of their decision upholds the raison d’être of an institution. Officeholders who are tasked with the decision about whether to declare a client’s funds as clean should actually be able to tell themselves and their fellow colleagues that their decision is based on reason.

Accountability, in the end, is not a technical problem. It is a cultural one. And no amount of regulation fixes the culture of an institution that has quietly learned that asking too many questions is bad for business.

Emanuela Ceva, Geneva RCCS

05 May 2026

Albania's appointment of Diella, an AI-generated “minister” associated with public procurement, has made global headlines. The official government profile presents Diella as Minister of State for Artificial Intelligence, while Reuters reported that Prime Minister Edi Rama introduced it as a way to make public tenders “100% free of corruption.” The promise is politically powerful: remove human discretion from procurement, and corruption will disappear. The promise is also too quick.

The question is not mainly empirical regarding whether AI can help anticorruption. It can. AI systems can detect anomalies, compare bids, flag suspicious patterns, and make procurement files more transparent. The harder question is ethical and concerns whether AI could and should replace officeholders’ direct engagement, characterized by reflective and critical albeit fallible exercise of public power. From the perspective developed at the Geneva Research Center for Corruption Studies (RCCS), the answer is a qualified “no.”

The RCCS studies corruption as a deficit in the accountability chain linking officeholders to their mandates. In Political Corruption: The Internal Enemy of Public Institutions, Maria Paola Ferretti and I argued that corruption is not only bribery or personal enrichment. It is an unaccountable use of the power of office: a use for an agenda whose rationale is unjustifiable by reference to the power mandate. This matters because corrupt institutional action can occur even when no one pockets money.

Seen from this perspective, Diella's “ministerial” framing is ethically risky. Public procurement is not a mechanical exercise. It requires judgment: officeholders must interpret eligibility criteria, assess trade-offs in case of disagreement, and generally act in ways aligned with their best and bona fide interpretation of the raison d’être of their institution. If the decisive choice is attributed to a system whose reasoning cannot be reconstructed into reasons that officeholders can own, institutional authorship becomes blurred. Officeholders may then say that “the computer decided.” That is not anticorruption. It is responsibility deflection.

This concern does not entail a rejection of AI. In Automating Anticorruption?, María Carolina Jiménez and I argued that machine-learning tools may support anticorruption only if they strengthen, rather than weaken, office accountability. Diella could therefore be designed as a sentinel, not as an adjudicator. It may flag risk indicators, identify unusual bidding patterns, generate structured comparisons, and help oversight bodies focus attention. But final decisions should remain with identifiable officeholders who exchange reasons, answer objections, and can be reviewed by other officeholders and by the public, thus claiming co-authorship of any institutional act.

Anticorruption is not achieved by evacuating human discretion from institutions. It is achieved by structuring discretion so that those who exercise public power can give and demand reasons for how that power is used. Only under those conditions AI can uphold institutional functioning as an accountable work in progress.