Soft law, soft behaviour and hard truths – speech at IBA Arb40 Symposium 2026

VK Rajah SC
Overseas Member, Duxton Hill Chambers London

Colleagues and friends, thank you for the invitation to speak.

I want to talk plainly about international arbitration, about where we have come from, where we stand today, and what we must confront if we are serious about restoring confidence in the system.

My core message is simple: we need to stop performing the rituals of a system that we say is efficient and ethical, and start building one that truly is. The fundamental issue is not a shortage of reforms, rules, or soft-law guidance; it is the absence of accountability that bites when actors behave poorly. Unless we are prepared to confront that truth, inefficiency, delay, and ethical drift will persist despite our best-intentioned tweaks at the margins. It is time to remove the window dressing, look honestly at what is inside, and rebuild a system in which accountability is not aspirational but central.

Naming our reality: from promise to performance

Half a century ago, international arbitration was a lean and pragmatic craft: flexible, bespoke, and closely aligned to the needs of commerce. Over time, it evolved into a sophisticated transnational framework, underpinned by treaties, model laws, institutional rules, and an expanding soft-law corpus. That evolution delivered something transformative: global enforceability and procedural autonomy that national courts could not consistently provide across borders. This progress was real, and it was hard-won.

However, there was, from the outset, a central paradox. Arbitration sought the assistance of the public authorities from which it wanted to free itself. Yet courts eventually crossed the Rubicon, from suspicion to support, and, in doing so, enabled the modern system we now take for granted. But success has bred complacency. We have come to rely on rituals and slogans such as efficiency, neutrality, user‑centricity, even as users increasingly encounter escalating costs, delay, opacity, unmanaged conflicts, and ethical standards that lack an enforcement backbone. It is time to name that reality.

Ethics and accountability, as presently constituted, are underpowered. Instruments such as the IBA Guidelines exert undeniable influence, but influence is not enforcement. Repeat appointments concentrate authority without corresponding consequences for lapses or failures. Popular arbitrators become structural bottlenecks. Sanctions are rare. In this ethical wilderness, confidence inevitably erodes, particularly among users outside the traditional centres, for whom the suspicion already exists that the game is played by and for insiders.

When tribunals do attempt to impose discipline, they are too often cautioned against the spectre of ‘due process paranoia’, notwithstanding the fact that jurisprudence at leading seats consistently affirms robust and decisive case management. These are not isolated irritants. They are symptoms of a deeper disease: weak accountability.

Stop performing, start building: accountability that bites

If self‑regulation is a myth without credible enforcement, what follows? The answer is not to abandon autonomy, but to anchor it in accountability with consequences that are visible, proportionate, and portable across institutions. That is how we turn slogans back into truths.

First, we should mandate transparent, published reasoning on challenges, removals, and conflicts. The object is not exposure, but discipline: a public corpus that constrains behaviour and educates the market, importing the court‑like virtues of transparency and reason‑giving without compromising confidentiality on the merits. Without principled visibility, norms drift and trust decays.

Second, we need binding, cross‑institutional ethical enforcement. Today’s guidelines are soft; tomorrow’s should have consequences. A compact among leading institutions and provider bodies could create reciprocal recognition of disciplinary outcomes for serious breaches, non‑disclosure of conflicts, undisclosed concentrations of repeat appointments, discovery abuse – administered by an independent panel observing due process. Soft law, when backed by hard consequences across platforms, will reshape incentives decisively.

Third, tribunals must be empowered and expected to manage cases decisively. Supervisory courts afford very considerable deference to robust case management. Institutions should therefore codify presumptions against late evidence absent good cause, constrain guerrilla tactics, and require early, tribunal‑led timetabling tied to cost consequences. This is not anti‑due process; it is due process that recognises time and money as variables of justice, rather than just collateral damage.

Fourth, we must open the pipeline and widen the circle. Publish diversity and repeat‑appointment metrics by case, not merely in aggregate. Expand list procedures and curated blind appointment pilots. Consider capping serial party appointments in tightly connected sectors for limited periods to mitigate concentration risk. Quality will assert itself, our duty is to ensure the market can actually see it.

Fifth, we must reconnect arbitral practice with the public law of commerce. International commercial courts from Singapore’s SICC to the DIFC are signalling strong user demand for accountable, visible law‑making alongside private adjudication. Properly configured, courts and arbitration are complements, not competitors. Institutions should partner with ICCs to align procedural best practice, cross‑reference decisions where appropriate, and stabilise expectations across public and private adjudicative tracks.

Why reforms keep missing the point

We have not lacked for reforms; we have lacked for accountability. We tinker with rules, issue fresh guidelines, and congratulate ourselves at conferences, yet inefficiency and ethical unease remain. That is because procedural reform without enforcement is theatre, not governance. An ethics regime that depends primarily on reputational sanction invites the very opacity that it purports to cure. Where consequences do not follow conduct across institutional boundaries, ‘soft law’ becomes an alibi for soft behaviour. That is why costs spiral, timelines slip, and confidence frays.

In recent years, the centre of gravity in standard‑setting has migrated towards providers and transnational professional bodies. They have delivered valuable coherence and technical refinement, but they are not a substitute for user accountability. The risk is a producer‑driven ecosystem: committees and conferences generating norms for constituencies who are not in the room. Legitimacy cannot be manufactured internally; it must be earned externally.

Users, meanwhile, are signalling their dissatisfaction with growing clarity. Persistent concerns over delay, cost, and opacity have not abated. The rise of international commercial courts is not a repudiation of arbitration; it is a demand signal – for transparency, for reasoned and visible guidance, and for judicial authority operating alongside private adjudication. We can disregard that signal, or we can integrate with it. The future legitimacy of the field turns on which course we choose.

Learning from other transnational systems

Arbitration is not unique in having to reconcile autonomy with accountability. In other transnational sectors as varied as shipping, sport, banking and telecommunications, legitimacy rests on more than technical rules; it rests on enforcement that is credible, cross‑border and visible enough to shape behaviour. Where those systems have matured, it is because accountability has been designed to travel: recognised across platforms and backed by shared standards. That is the direction that arbitration must now take, by coupling soft‑law norms with consequences that bind across institutions. The lesson is clear: when incentives change, conduct changes; when enforcement is portable, trust becomes systemic, rather than episodic. That is the path we must take.

The parallel to arbitration’s current condition is instructive. We have strong technical architecture such as treaties, model laws and institutional rules but we have allowed accountability to remain largely informal, reputational, and siloed. In practice, that leaves users facing delays, spiralling costs, and conflicts with insufficient recourse, and leaves tribunals wary of managing cases decisively for fear of imagined enforcement risk. We should take the opposite lesson: courts from the United States to China today set a high bar for interference; decisive, transparent case management is not merely safe, it is expected.

Putting users genuinely at the centre

We say users should be central; let us make that literal. Large cross‑border users can insist on ‘user charters’ as a condition of choosing seats and institutions. Such charters might include commitments to published reasoning on challenges, enforceable ethical standards through reciprocal disciplinary recognition, transparency in diversity and repeat‑appointment metrics, and principled governance for the use of AI in proceedings.

That is not idealism; it is values‑based realism: accept the world as it is and act with honesty to build the institutions that we claim to believe in. Users especially from the Global South, and among newer exporters of capital, have the collective capacity to recalibrate the system’s centre of gravity by aligning contracts, panel selections, and institutional choices with standards that reward accountability and integrity.

Re-centring users also requires us to confront representation. The map of commerce and capital has changed decisively. Many jurisdictions, east and west of Suez, now export both capital and legal capacity, yet they remain under‑represented on committees, rosters, and governing boards, often as a result of legacy selection practices. If our field is to retain its legitimacy, leadership must reflect where commerce now resides, not where it was two generations ago.

AI: accelerating accountability, not opacity

Artificial intelligence will accelerate every aspect of dispute resolution: the generation of analysis, evidence, conflict checks, procedural drafting, and beyond. Used well, it can strengthen accountability by illuminating patterns of repeat appointments, mapping counsel–arbitrator networks and identifying cost or time outliers. Used poorly, it risks deepening opacity. The choice before us is therefore not whether to adopt AI, but how.

We should choose transparency. Material reliance on AI in proceedings should be disclosed, subject to protocols that safeguard confidentiality and privilege. Institutions should deploy analytical tools to identify concentration risks and publish aggregated insights that nudge better behaviour without compromising case confidentiality. Judges and arbitrators should be trained together, cultivating a shared literacy in AI-assisted evidence, bias risks, and explainability.

Finally, soft and hard law must move in concert. Iterative, UNCITRAL‑style frameworks addressing AI‑related evidence, disclosure obligations and cybersecurity can provide a common baseline, progressively harmonised across arbitral institutions. In this way, technology becomes not a source of mistrust, but an instrument for strengthening confidence in the system itself.

Taking the sign down

Let me end where I began. Václav Havel once wrote of the greengrocer who displays a sign that he does not believe, sustaining a system through ritual rather than conviction. Mark Carney recently revived that parable to challenge us to stop invoking a rules‑based order that no longer operates as advertised and instead to build institutions that do what they claim. Our field now confronts a similar moment of reckoning. We speak readily of efficiency, neutrality, autonomy, and accountability. It is time to make those words true again.

That task is neither abstract nor elusive. It begins with ethics that are enforceable and portable across institutions. With transparent and reasoned decisions on conflicts and challenges, published in anonymised but instructive form. With firm case management supported by the principled deference of supervisory courts. With leadership and representation that reflect where commerce now resides. And with a smart integration with international commercial courts and AI‑ready standards that accelerate accountability rather than deepen opacity.

Power, in any system, tends to entrench itself. But you, especially the next generation, possess something just as consequential: the capacity to stop pretending. To insist on accountability that bites. To choose seats and institutions with backbone. To build strength at home, and to act together across borders. If we do this, we can finally take the sign out of the window by aligning what we say with what we do, and, in a world of shifting capital and accelerating technology, preserve commercial justice that is not only private and efficient, but principled and legitimate.

That is the work of our time. Let us begin.