The new Arbitration Act 2025: Necessary modernization or missed opportunity?

Background

The UK has long stood at the forefront of international arbitration. It was home to one of the first arbitration statutes in the world, John Locke’s Arbitration Act 1698, with the Arbitration Act 1996 being instrumental in securing London’s status as a leading seat for international arbitration. Indeed, in the 2025 International Arbitration Survey, London was ranked as the top choice seat for respondents, arbitrators and counsel.  The 1996 Act is considered one of the leading statutory frameworks for international arbitration and, following a wide-ranging Law Commission consultation, the 2025 Act is intended to build on its success.

Key Reforms to the 2025 Act

1. Default rule for governing law of arbitration agreement

Section 1 of the 2025 Act (which creates Section 6A of the 1996 Act) introduces a new default rule: if an arbitration agreement does not expressly specify the law which governs it, it will be governed by the law of the seat, not the law governing the main contract. This brings English law into concord with a number of other major arbitration centers including France and Sweden, and marks a departure from the Supreme Court’s approach in Enka v Chubb [2020] UKSC 38 which (in effect) established a default rule that the governing law of the main contract, not the seat, applies to the arbitration agreement if the arbitration agreement is otherwise silent. 

This reform has generated the most discussion, with practitioners putting forward various suggestions, including adopting the Enka v Chubb approach espoused by the Supreme Court and limiting the default rule only to arbitrations seated in England and Wales. 

The clarity introduced by Section 6A is to be welcomed after years of debate and expensive litigation. There are, however, some uncertainties as to what an “express” choice of governing law for the arbitration agreement means and therefore when the new default rule will apply in practice:

• In Kabab-Ji SAL (Lebanon) v Kout Food Group (Kuwait) [2021] UKSC 48, the UK Supreme Court – in a case concerning the governing law of an arbitration agreement – found that a term derived from implication is itself an express choice because it reflects the parties’ choice. Applying this reasoning, it is conceivable that a Court could find that, whilst there is no governing law clause included in the arbitration agreement, the parties have nonetheless by implication made a choice of governing law and that consequently the default rule does not apply.

• In UniCredit Bank GmbH v RusChemAlliance LLC [2024] UKSC 30, another arbitration case which reached the UK Supreme Court, it was found that the words “non-contractual and other obligations” (which are commonly found in governing law provisions) may be sufficient to encompass an arbitration agreement and thereby constitute an express choice of law governing the arbitration agreement. This is another situation where it is possible that the default rule will not apply even though there is no “written” governing law clause.

Although the Law Commission only considered these uncertainties in passing, it appears to envisage that the default rule should apply to both the above scenarios. The Courts will need to clarify whether this is the case. From a practical perspective, it continues to be good practice to avoid all uncertainty by including an express clause setting out the choice of governing law of the arbitration agreement (particularly where this differs from the law of the seat).

2. New power of summary dismissal

Section 7 of the 2025 Act (which creates Section 39A of the 1996 Act) confirms the power of a tribunal to issue awards on a summary basis upon an application made by a party (a point that had previously been the subject of some debate) unless the parties agree otherwise. This power can be used if – after giving each party a reasonable opportunity to put forward its case – the tribunal considers that a party has “no real prospect of success” on a relevant issue. This standard reflects that applied in English civil procedure and aligns with the rules of many leading institutional rules (such as the LCIA and ICC rules). 

The Law Commission’s aims are twofold: (1) to prevent parties with weak cases from abusing the arbitral process, and (2) to give arbitrators clear reassurance that making a summary award will not breach their duty to give each party a reasonable opportunity to present their case or give rise to a ground for challenging the award. This clarification addresses one of the perceived pitfalls of arbitration, that is, that tribunals do not have the toolkit to deal efficiently with meritless claims.

3. Revised framework for jurisdiction challenges under Section 67

Section 11 of the 2025 Act (which amends Section 67 of the 1996 Act) introduces a new framework for when a party challenges an award for lack of substantive jurisdiction.

Under the reforms:

• A party cannot raise a ground for objection before the Court unless it raised it before the tribunal (save where the party did not know and could not with reasonable diligence have discovered the ground during the arbitration);
• A party cannot rely on evidence before the Court unless it was put before the tribunal (save where the party could not, with reasonable diligence, have done so); and
• “Evidence that was heard by the tribunal must not be re-heard by the court”.

This amendment limits the Court’s ability to rehear evidence already presented to the tribunal and restricts parties from introducing new grounds or evidence at the award stage. It is a material change to the 1996 Act, which provided that jurisdictional challenges under Section 67 were effectively a de novo review by the Court and often included new arguments and evidence as well as a rehearsal of those points determined by the tribunal.

The aim of this reform is to promote finality and efficiency in English law arbitrations by reducing costs and avoiding unnecessary delays. It will ensure procedural fairness by preventing parties – who have already had their case examined during the arbitration phase – from attempting to refine or reargue their position in Court. The consequence is that the number of Section 67 challenges – already rare – should be reduced.

4. Clarification of Court powers in support of emergency arbitrators

Section 8 of the 2025 Act (which creates Section 41A of the 1996 Act) makes express provision for the appointment of emergency arbitrators where the parties have agreed to arbitration rules that provide for such appointments. It also grants emergency arbitrators the authority to issue peremptory orders, ensuring that these orders can be enforced by the Courts in the same way as those issued by non-emergency arbitrators. The 1996 Act was silent on emergency arbitration, as the concept only emerged after its enactment. Consequently, there was uncertainty about the enforceability of emergency orders and this reform provides useful legal clarity. It reassures parties that pre-emptory orders issued in urgent situations are enforceable and reinforces England and Wales as a reliable and effective seat for resolving time-sensitive disputes. 

5. Codification of arbitrators’ duty of disclosure 

Section 2 of the 2025 Act (which creates Section 23A of the 1996 Act) introduces a statutory duty requiring arbitrators to disclose any circumstances that could reasonably give rise to justifiable doubts about their impartiality, viewed through the eyes of the parties. Common examples of circumstances that require disclosure include prior involvement or a financial interest in the dispute, or relationships with parties, their representatives or their witnesses.  

This reform effectively codifies the disclosure rule formulated by the UK Supreme Court in Halliburton v Chubb [2020] UKSC 48 and aligns with international best practice (for example, the UNCITRAL Model Law, the ICC Rules, the LCIA Rules and the IBA Rules on Conflicts of Interest). The 2025 Act does, however, go further insofar as it requires an arbitrator to disclose any relevant circumstances of which they “ought reasonably to be aware,” and therefore imposes an obligation on arbitrators to conduct diligent enquires into the existence of relevant circumstances before and after accepting appointments.

It is unlikely that Section 2 will lead to a material change in established good practice. Arbitrators already typically make relevant disclosures in a timely way (particularly if there are institutions administering their appointment and the arbitrations). However, its addition to the Act highlights the importance of making adequate disclosures and gives arbitrators a clear statutory framework to follow. 

6. Arbitrator immunity 

Sections 3 and 4 of the 2025 Act (which amend Sections 24, 25 and 29 of the 1996 Act) enhance protections for arbitrators against liability arising from removal or resignation, reversing a string of decisions suggesting that arbitrators could be liable for costs associated with their removal. In summary: 

• Section 3 of the 2025 Act clarifies that an arbitrator will only be liable for costs associated with their removal where it can be shown that they acted or omitted to act in bad faith.
• Section 4 of the 2025 Act provides that an arbitrator will only be liable for costs associated with their resignation in circumstances where their resignation was, in all the circumstances, unreasonable. 

The Law Commission proposed these reforms to ensure arbitrators can make decisions without fear of accruing personal liability. Notably, the 2025 Act draws a clear distinction between the resignation and removal of arbitrators. The introduction of a higher threshold for liability in cases of removal (that is, it needs to be shown that the arbitrator acted in bad faith compared to the “reasonableness” test for resignations) is a proportionate safeguard, given the potential for such claims to undermine confidence in the arbitral process. 

Proposed reforms not adopted in the Act

Broadly speaking, the 2025 Act makes only modest changes to the 1996 Act and, even then, not all the recommendations made by the Law Commission and the arbitration community have been adopted in the Act. 

Duty of confidentiality 

The 1996 Act has been criticized for not providing a clear statutory duty of confidentiality, particularly given the uncertainty that remains over its source, scope and extent in English arbitration law. 

In practice, London-seated arbitrations are deemed to be private and confidential, with the duty of confidentiality arising by virtue of a term implied by operation of law (unless there are express provisions addressing the point). Some institutional rules, for example, Article 30 of the LCIA Rules, address confidentiality directly by codifying the English common law position that is, in effect, that all materials created for the arbitration are to be kept confidential unless disclosure is required by law. The Tribunal may also deal with confidentiality in its procedural orders.

The 2025 Act was an opportunity to introduce a consistent framework for confidentiality in arbitrations seated in England and Wales. However, in its report, the Law Commission concluded that a statutory “one-size-fits-all” approach to confidentiality would not be appropriate. Instead, it endorsed the current, arguably more flexible position – under which parties can choose from a range of confidentiality mechanisms – as better suited to the diversity of arbitral proceedings. This approach allows the parties to retain greater control over the conduct of their dispute and may also promote more transparency in arbitration – a principle for which there is growing support in some areas of the arbitral community. The counterargument, of course, is that parties generally expect arbitrations to be confidential – it is often touted as a key benefit of the process – and the current position is something of patchwork that leaves parties wondering what information they can disclose about ongoing arbitrations to, for example, their auditors or the market and what must be kept confidential.

The Law Commission acknowledged that the “balance between confidentiality and transparency is still a matter of debate” and that debate will continue.

Third-party funding

Third-party funding (TPF) – where a third-party covers a party’s legal costs in exchange for reimbursement and a share of any recovered sums – is increasingly common in international arbitration. Concerns have emerged around issues including disclosure and conflicts of interest, particularly where arbitrators may have ties to funders. 

In England and Wales, further complications have arisen from the Supreme Court’s 2023 decision in R (on the application of PACCAR Inc) v Competition Appeal Tribunal [2023] UKSC 28. The Supreme Court held that, where a TPF seeks a percentage of any damages recovered, the funding agreement qualifies as a damages-based agreement (DBA) under the Courts and Legal Services Act 1990 (CLSA 1990). DBAs are unenforceable unless they comply with the Damages-Based Agreements Regulation 2013. While this ruling was made in the context of litigation, it cast uncertainty over arbitration funding. The ambiguous wording of the CLSA 1990 raises the possibility that Courts could find similar TPF arrangements in arbitration, particularly those governed by English law, unenforceable. 

The Law Commission did not address these issues in its report, nor did it consider TPF more broadly. To some extent, this omission is surprising. Introducing a statutory requirement to disclose the existence of TPF would have aligned English law with the position in Hong Kong and Singapore, leading institutional rules (such as the SIAC, HKIAC and ICC Rules) and the IBA Guidelines on Conflicts of Interest.

That said, the Civil Justice Council’s (CJC) report on TPF is expected in the summer of 2025. This may lead to legislation to clarify the extent of any obligation to disclose funding arrangements and the enforceability of funding agreements in the context of arbitration. 

Regulation of technology

The 1996 Act does not explicitly address the use of technology and modern working practices in commercial arbitration. Under Section 34 of the 1996 Act, arbitral tribunals have broad discretion to determine all procedural matters (subject to party agreement). In light of this, the Law Commission concluded that no reform is needed to address technology in arbitration as Section 34 is sufficiently flexible to allow arbitrators to adopt modern technology. 

Some institutional rules have been more prescriptive, introducing provisions that govern the use of technology including the remote examination of witnesses, remote hearings and electronic documentation. As artificial intelligence and other technological innovations become more commonplace in arbitration, a more formal regulatory framework may be required to ensure consistency, fairness, and efficiency across proceedings.

Arbitrator independence

The Law Commission chose not to introduce an express statutory duty of independence for arbitrators. Had such a duty been adopted, it would have required arbitrators to have no prior connection to the arbitrating parties or the dispute – going beyond the existing requirement of impartiality, which focuses on neutrality and a lack of bias. Although similar duties exist in the UNCITRAL Model law and the ICC Rules, and in other jurisdictions, for example, Paris, Singapore and Hong Kong, the Law Commission concluded that impartiality – and the appearance of impartiality – is a more appropriate standard in the context of English arbitration. It considered that focusing on actual and perceived bias offers a more practical and meaningful safeguard than a more rigid and objective independence requirement.   

The Law Commission also highlighted the difficulty of achieving full independence in practice, especially in specialized industries where the pool of suitably qualified arbitrators is small. In these sectors, arbitrators often have prior professional relationships with parties, counsel or other arbitrators – indeed, this is why they will have been appointed – and freedom of choice of arbitrator is often seen as a key advantage of arbitration over the Courts. Imposing a strict independence rule could limit party autonomy and complicate appointments in niche markets. 

Discrimination 

In its report on the 2025 Act, the Law Commission did not recommend any reform to the 1996 Act to explicitly address discrimination. The Commission noted that the arbitrator's duty to act fairly and impartially already encompasses an obligation to avoid discriminatory conduct. Where concerns arise over an arbitrator’s impartiality – including any stemming from discriminatory behavior – there are mechanisms for removal. Aspects of the arbitral process are also already subject to anti-discrimination protections under the Equality Act 2010. That said, the Commission identified a gap in the current position: arbitration agreements can, in some cases, include discriminatory criteria for appointing arbitrators – for example, requiring that an arbitrator be of a specific nationality. While acknowledging this as an area of concern, the Commission concluded that banning such clauses could give rise to complex enforcement issues and lead to satellite litigation and decided not to recommend reform on this point.   
Although the 2025 Act does not introduce any explicit prohibition on discrimination in arbitration, there are various initiatives in the sector to promote greater diversity within the field. 

Impact of the 2025 Act

The 1996 Act required some degree of modernization, particularly because competitor seats, for example, Singapore, Switzerland and Germany, have recently taken steps to modernize their arbitration laws.

However, the fact that the 2025 Act’s amendments are largely incremental or clarificatory is a testament to the success of the 1996 Act and the strength of London's international arbitration market, comprising 5,000 arbitrations annually and said to contribute over £2.5 billion to the UK's economy. The 2025 Act has rightly been described as evolution, not revolution. While a case can be made that the 2025 Act, being the first reform in nearly 30 years, could have been more ambitious, the changes it introduces are progressive while retaining the essence of what made the 1996 Act so effective. 

The authors would like to thank Max Sharp for his assistance in the preparation of this article.