Framework at a glance
A major commercial dispute should be managed as a continuing commercial decision, not only as a sequence of legal steps. The client’s objective and appetite for risk provide the reference point. The strategy remains sound only while the assumptions supporting it continue to hold.
The compass for the dispute. What is the client trying to achieve, and which trade-offs are acceptable?
Legal analysis · Economics · Recoverability · Intelligence · Behaviour · External developments
Integrates the relevant inputs around the actual choice. Makes assumptions, sensitivities and information gaps visible.
The course selected on the information available, together with the assumptions that must remain sufficiently true for that course to make sense.
Maintains integration across workstreams, monitors material change, identifies inconsistencies and brings developments back into the decision process.
Technology can support monitoring and analysis. The client remains responsible for objectives, risk appetite, trade-offs and the final decision. Counsel remains responsible for legal advice and proceedings.
From legal merits to commercial value
1. Why winning is the wrong starting point
A board considering a major commercial dispute may have strong legal advice, a detailed financial model, an assessment of the counterparty’s assets, litigation analytics and specialist intelligence. Each may be rigorous. Together, they may still leave the board without a satisfactory answer to the question that matters.
What should we do?
The difficulty is not simply one of information. Major disputes already generate more analysis than most decision-makers can absorb. The harder problem is that different advisers examine different parts of the dispute, often at different times and through different assumptions.
Counsel assesses the legal merits. Financial analysis estimates cost and value. Investigators examine ownership, assets, relationships and behaviour. Management considers capital, commercial priorities and reputational consequences. Each perspective may be necessary. None, on its own, determines the best course for the client.1
This report advances a straightforward proposition: a major commercial dispute should be managed as a continuously reassessed commercial decision, with clear ownership of the assumptions connecting legal strategy to the client’s actual objective.
Legal merits remain fundamental. They do not establish whether a claim is worth pursuing, whether a judgment is likely to produce payment, or whether another course would better serve the client’s objective.2 Rising costs, asset movements, deteriorating liquidity, new evidence, regulatory developments or a change in the counterparty’s incentives can make an initially rational strategy progressively less attractive.3
Decision Intelligence is the disciplined integration of legal analysis, economics, recoverability, intelligence, behaviour and external developments around a defined commercial objective. Its purpose is not simply to produce more analysis, but to determine what information matters to the decision, which assumptions the strategy depends on, and when those assumptions need to be reconsidered.
It should not be confused with decision automation. Models, analytics and AI can process evidence, test scenarios, monitor developments, identify inconsistencies and recognise when an assumption may no longer hold. They cannot determine what the client should seek to achieve, what level of risk it should accept, or which trade-offs are commercially tolerable.
What is often missing is ownership of the integration between workstreams. The report refers to that continuing mechanism as the Decision Function. It does not replace the client, General Counsel or lead counsel. It may be carried by an experienced individual, a team, technology, or a combination of all three.
That is the difference between winning the case and winning the dispute.
2. A strong claim can still be a poor commercial decision
A strong case may still be commercially unattractive if it is expensive, slow, difficult to enforce or disproportionate to the value ultimately recoverable. Conversely, a less certain claim may justify decisive action if delay, asset movement or strategic leverage makes waiting more costly.
Legal prospects measure the strength of the case. They do not measure the value of pursuing it.
Probability matters alongside quantum. Cost, duration, funding, cash flow and recoverability affect what the claim is worth to the client. Litigation must be compared not only with abandonment, but with settlement, further investigation, protective measures, external finance, negotiation or delay.
Burford Capital’s London Disputes Report 2026 found that 67% of respondents had seen strong legal claims go unpursued because of cost or risk concerns. Seventy-three per cent reported that boards had become more sophisticated in viewing disputes as financial assets. Eighty-five per cent agreed that risk-transfer tools such as legal finance improve litigation decision-making.4
Not every claim should be treated as a conventional investment. A dispute may concern control, reputation, precedent, deterrence or a relationship whose importance cannot be reduced to expected return. Even so, the decision to pursue a claim must survive a business case as well as a legal assessment.5
This is the governing question. Legal merits define what may be achievable. Economic analysis tests what the available paths may cost and produce. The client’s objective, priorities and appetite for risk determine which outcome is worth pursuing and which route is acceptable.
| Question | Principal focus | What it cannot determine alone |
|---|---|---|
| Can we win? | Law, evidence, procedure and remedies | Whether pursuing the claim serves the client’s wider interests |
| Is it worth pursuing? | Value, probability, cost, duration, funding and recovery | Which outcome the client should prioritise |
| What are we trying to achieve? | The commercial objective and the available routes to it | Whether the assumptions supporting the chosen route will continue to hold |
There is no universal definition of success outside the client’s objective.
A favourable assessment of liability is the beginning of the decision, not its conclusion. A claimant may establish liability and prevail on the law and the facts. The commercial value of that result depends on what happens next.
From judgment to recovery
3. Judgment is not cash
A judgment resolves a legal question and creates an enforceable right. It does not guarantee that the claimant will be paid. Considerable attention may be given to the prospects of establishing liability while recoverability is left until much later. By then, assets can have moved, ownership structures can have changed, and a defendant that appeared substantial may present a much more difficult enforcement profile.
The first concerns financial capacity. The second concerns incentives and behaviour: means without a reason to pay voluntarily. The third concerns legal and practical reach. Assets can exist without being easy to identify, preserve or enforce against. They may sit behind companies, trusts or nominees, or in jurisdictions where recognition and execution are slow, expensive or uncertain.6
Apparent wealth is not recoverable value.
In a cross-border dispute, the claimant needs to understand who owns or controls the relevant assets, where they are situated, what other claims may rank ahead, and whether enforcement is realistically achievable. Formal ownership does not always answer those questions.7
An early recoverability assessment is not meant to produce certainty. It is meant to establish whether recovery risk is material enough to affect the client’s course: whether the claim is pursued, the forum and defendants, the urgency of protective measures, the amount the claimant is prepared to spend, the price of funding, and the timing and level of settlement.8
Early analysis can also reveal leverage that has little to do with eventual execution. A financing event, transaction, ownership dispute, regulatory vulnerability or commercial dependency can alter willingness to engage. The existence of wealth is not enough. The claimant needs a realistic path from legal entitlement to realised value.
4. The economics of the dispute
Once a dispute is approached as a commercial decision, its value cannot be measured simply by the amount claimed or by the prospects of establishing liability. The relevant comparison is between the economic consequences of the courses available: litigation, settlement, further investigation, protective action, external funding, delay, or a negotiated commercial solution.
Expected dispute value ≈ (probability of success × value of outcome × probability of recovery), adjusted for cost, duration, disruption and strategic risk.
The framework is useful less for the number it produces than for what it exposes: which assumptions are carrying the decision, how sensitive the outcome is to changes in those assumptions, and where further information could genuinely affect the preferred course.
Herbert Smith Freehills Kramer has described models that assign probabilities to legal and factual pathways, discount future costs and outcomes, and test how the result changes when important assumptions move. In one published example, more than 20,000 possible routes through the issues were modelled.9
Its limitation is that the dispute does not stand still. Behaviour adds another dimension. Burford’s 2026 research reported that 60% of respondents agreed that settlement decisions are often driven by management fatigue rather than case strength.10 The counterparty’s willingness to settle, resist, delay or escalate cannot always be inferred from the legal merits.
A dispute does not stand still. Neither should the assumptions supporting its strategy.
From prediction to judgment
5. Prediction is not decision
Litigation analytics can identify patterns across courts, judges, counsel, parties, duration, procedural outcomes, damages and remedies.11 A widely cited study by Katz, Bommarito and Blackman predicted more than 28,000 United States Supreme Court case outcomes and about 240,000 individual justice votes, reporting 70.2% accuracy at case level and 71.9% at justice-vote level.12
Suppose the available analysis suggests a 68% prospect of success. That estimate may improve the assessment of the legal case. It does not determine whether the client should issue proceedings now, investigate further, seek protective measures, negotiate, accept an offer or wait.
Predictive accuracy is not the same as decision value. A reliable forecast on one legal issue may have little effect on the available courses. An uncertain fact about an asset movement, financing requirement or impending transaction could be decisive if it changes leverage or the cost of waiting.
The information that matters most is the information that can change the decision.
Gartner describes Decision Intelligence platforms as combining decision modelling, analytics and AI to augment or automate decision-making.13 For commercial disputes, the important development is less the prospect of automating judgment than the ability to organise a wider range of information around an actual choice. Technology can support the Decision Function. It cannot determine what the client should value.
From a fixed strategy to continuous reassessment
6. Reassessing strategy as the dispute changes
Commercial disputes develop around legal milestones: assessment, issue, pleadings, evidence, applications, settlement windows, judgment, enforcement. These matter because the process imposes them. They do not necessarily coincide with the moments when the client should reconsider its strategy.
An asset can move while the procedural timetable remains unchanged. Liquidity can deteriorate between hearings. A financing requirement, regulatory intervention or proposed transaction can shift bargaining positions without moving the case into a new legal phase.
A major dispute therefore runs on two tracks. The legal process is governed by procedure and the development of the case. The decision process is governed by the client’s objective, the assumptions supporting the current strategy, and changes in the circumstances on which those assumptions depend.
A legal milestone is created by the formal conduct of the case. A decision point arises when new information is significant enough to call an assumption into question. Not every development should reopen the decision. Material change can be a single event or a gradual shift: costs beyond expectation, weakening liquidity, a series of small asset movements.
- Define the objective. What the client is trying to achieve, and which trade-offs it will accept.
- Select the current course. The action that best serves that objective on the information available.
- Identify the assumptions that matter. The legal, financial, recovery or behavioural judgments that support the choice.
- Identify the conditions that would challenge them. Discrete events, agreed thresholds, or gradual changes.
- Reassess what has changed. Revisit the affected assumptions rather than reopening every aspect of the dispute.
- Decide whether the strategy still holds. Continue where its basis remains sound. Change course where the balance has shifted.
Technology can track agreed assumptions, connect information from different sources and surface changes that warrant attention.14 Its value is not to decide whether the strategy should change. It is to reduce the risk that a relevant change passes unnoticed. Recognising the change early gives the client a choice. It does not necessarily mean acting immediately.
From separate workstreams to an integrated decision
7. The Decision Function
Excellent advisers do not, by themselves, create an integrated decision process. Counsel may assess settlement against an assumed level of recoverability. A financial model may rely on a probability that no longer reflects counsel’s view. Investigators may identify a change in ownership without knowing whether it affects an asset relevant to enforcement. Management may become less willing to absorb delay without that change entering the strategy.
The gap is not necessarily the absence of another adviser. It is the absence of a function that connects these inputs and keeps the basis of the strategy coherent. The client remains responsible for the objective, risk appetite and final decision. Lead counsel remains responsible for legal advice and the conduct of the case. The Decision Function sits across those workstreams. The issue is not who holds a title. It is whether the function is being performed.
- Clarify the decision. “Should we litigate?” may be too narrow. The real choice may involve negotiation, investigation, protection, funding, more than one forum, or a commercial exit.
- Identify the assumptions that matter. Understand where the decision is most exposed to uncertainty, and who owns each assessment.
- Reconcile the workstreams. Test whether legal opinions, asset reports and models remain compatible.15
- Identify what is missing. More information is not always better information. The relevant question is whether it has decision value.
- Maintain visibility over material change. Some developments arrive as events. Others develop gradually.
- Challenge the recommendation. What would have to be true for another option to be preferable? Where do the principal risks sit?
Its value increases with multiple jurisdictions, substantial capital at risk, uncertain recoverability, several advisory teams, or an objective that extends beyond judgment. The principal risk is duplication. Technology can support the function. It does not change the allocation of responsibility.
When the legal case stays strong but the commercial decision changes
Illustrative composite
The following scenario is illustrative. It combines recurring features of complex commercial disputes and publicly observable cases. It does not describe a single K-Partners matter.
A company begins a substantial cross-border claim against a privately controlled group. Counsel considers the legal case strong. The amount at stake is significant and the defendant appears to have sufficient assets to meet a judgment. The client wants maximum recovery and is prepared to fund proceedings through judgment if necessary.
That strategy rests on assumptions that do not form part of the merits analysis: that the defendant will remain capable of paying, that important assets will remain accessible, that the proceedings can be funded without undue pressure, and that time broadly favours the stronger legal position.
As the dispute develops, the legal case remains largely unchanged. Corporate filings show changes in ownership. Intelligence indicates that a refinancing or transaction is becoming important to the controlling shareholders. Value is reorganised across the group. The claimant’s management becomes less willing to accept several further years of cost and uncertainty.
Counsel may still be confident about liability. The financial model may still show a substantial expected value. Viewed separately, neither analysis necessarily produces a reason to alter course. The significance appears when the inputs are considered together. A settlement that would previously have appeared premature may now produce greater realised value than a larger judgment several years later.
The legal strategy may still be right. The commercial decision may have changed.
A Decision Intelligence approach would bring those developments back into the same analysis. That could lead to further investigation, protective measures, a reassessment of recovery, and a more deliberate approach to negotiation while the refinancing gives the counterparty a reason to resolve the dispute. Another client with different priorities might continue to judgment.
The case may still be legally the same case. Commercially, it may no longer be the same decision.
8. Legal strategy is not dispute strategy
Legal strategy determines how best to use the law: which claims, against whom, in which forum, on what evidence, through which procedural steps. Dispute strategy asks how those legal options should be used to achieve the client’s commercial objective.
Counsel may conclude that proceedings should be issued immediately. The wider strategy may favour a further period of investigation if that work could identify assets or improve the negotiating position. A strong interlocutory application may still add cost, delay the wider strategy, or reduce the prospect of settlement.
A conclusion that the client has “good prospects” may be accurate and still incomplete. The client needs to know which assumptions matter most, what evidence would change counsel’s view, and which apparent legal victories would have limited effect on the wider objective. Herbert Smith Freehills Kramer has described software-enabled risk modelling to make legal risk examinable at board level.16 The value lies less in a single number than in making the reasoning and the range of outcomes visible.
The Decision Function does not reinterpret legal advice. Where a legal assumption is unclear or inconsistent with another workstream, the issue should be returned to counsel. The General Counsel often sits where legal advice becomes a business decision.17 In some organisations the GC will carry much of the Decision Function. In others, the scale of the dispute will require broader support.
For every significant legal recommendation, the client should be able to answer three questions. What does this improve in the legal case? What does it change in the commercial position? Does that change justify the cost, risk and likely response?
9. Beyond winning
Major commercial disputes rarely suffer from a shortage of expertise. Technology will continue to increase the volume and speed of that work.18 The harder problem is ensuring that expertise continues to serve the decision the client actually has to make.
A legal recommendation can remain sound while becoming less attractive commercially. A model can be rigorous while relying on assumptions that no longer reflect recovery. New intelligence can change incentives without entering the wider strategy. The client’s objective remains the central reference point.
The Decision Function does not replace counsel, management or specialist advisers. It maintains coherence between their inputs and makes clear when the basis of the current strategy has changed enough to warrant reconsideration. AI can help preserve continuity across a long-running dispute and reduce the risk that an important development is missed.19
Technology can expand the field of view. Judgment remains human.
Technology does not decide what the client should value, how much uncertainty it should tolerate, or which trade-offs are justified. Its value is that it can improve the conditions in which judgment is exercised.
Going beyond winning does not diminish legal victory. In many disputes, winning the case will be essential. Legal success has to be understood in the context of what it delivers, what it costs to obtain, and whether it remains the best route to the client’s objective.
The measure of a sound dispute strategy is not simply whether it was rational when adopted. It is whether it remained aligned with the client’s objective as the dispute itself changed.
Notes
- EY, “2025 General Counsel Study”, June 2025; FTI Technology and Relativity, The General Counsel Report 2025.
- Burford Capital, London Disputes Report 2026, 7 July 2026.
- Stolen Asset Recovery Initiative, Asset Recovery Handbook, 2nd ed., World Bank and UNODC, 2020; Financial Action Task Force, “Beneficial Ownership”.
- Burford Capital, London Disputes Report 2026. 67% had seen strong claims go unpursued because of cost or risk; 73% reported greater board sophistication in viewing disputes as financial assets; 85% agreed that risk-transfer tools such as legal finance improve litigation decision-making.
- EY, “2025 General Counsel Study”, June 2025; FTI Technology and Relativity, The General Counsel Report 2025.
- Stolen Asset Recovery Initiative, Asset Recovery Handbook, 2nd ed., 2020.
- FATF, Guidance on Beneficial Ownership of Legal Persons, 10 March 2023; Guidance on Beneficial Ownership and Transparency of Legal Arrangements, 11 March 2024.
- Stolen Asset Recovery Initiative, Asset Recovery Handbook, 2nd ed., 2020.
- Herbert Smith Freehills Kramer, “Using Decision Analysis to Evaluate a Settlement Offer”; “Decision Analysis — Putting legal risk in the language of the boardroom”, 15 March 2023.
- Burford Capital, London Disputes Report 2026. 60% agreed settlement decisions are often driven by management fatigue rather than case strength.
- Lex Machina, Legal Analytics API overview. A product description, not independent validation of predictive accuracy.
- Katz, Bommarito and Blackman, PLOS ONE, 12 April 2017. Historical US Supreme Court dataset: 70.2% case-level, 71.9% justice-vote accuracy.
- Gartner, Magic Quadrant for Decision Intelligence Platforms, 26 January 2026.
- Freshfields, “From Siloed Tools to Seamless Workflows”, 21 July 2026; Stanford Law School, liftlab, 15 September 2025.
- Herbert Smith Freehills Kramer, decision analysis publications, 2023 and settlement-offer analysis.
- Herbert Smith Freehills Kramer, software-enabled decision analysis service, 24 January 2019.
- EY, “2025 General Counsel Study”, June 2025; FTI Technology and Relativity, The General Counsel Report 2025.
- Lex Machina; Freshfields, 21 July 2026; Stanford CodeX.
- Freshfields, 21 July 2026; Stanford Law School, liftlab; Gartner, 26 January 2026.