Civil Law and UAE: Transition from Narrative Justice to Dataset Justice

1. Introduction

The expression “transition from narrative justice to dataset justice” is best understood as a conceptual description of how civil adjudication may evolve in a technologically advanced UAE legal system.

Traditional civil adjudication is largely narrative-based. Parties present:

• pleadings;
• witness statements;
• contracts;
• correspondence;
• expert reports;
• documentary evidence;
• oral explanations;
• competing factual stories.

The judge then determines which facts are legally established and applies the relevant legal rules.

Dataset justice does not mean that an algorithm replaces the judge. It describes a system in which increasingly large and structured bodies of information—electronic records, transaction histories, databases, blockchain records, metadata, CCTV, financial data, AI-generated material and statistical information—become central to fact-finding, case management and legal analysis.

This distinction is particularly relevant in the UAE because the legal system is undergoing simultaneous digital transformation and legislative modernisation. The new Federal Decree-Law No. 25 of 2025 promulgating the Civil Transactions Law entered into force on 1 June 2026, repealing the 1985 Civil Transactions Law.

At the same time, the DIFC Courts have established specialised infrastructure for disputes involving big data, blockchain, artificial intelligence, cloud services, digital assets and automated dispute resolution.

2. Meaning of Narrative Justice

Narrative justice

Narrative justice is a model in which the court reconstructs events from human accounts and documentary evidence.

For example:

“The defendant promised to deliver the goods, repeatedly assured the claimant that delivery would occur, but later refused to perform.”

The court must reconstruct:

1. what was agreed;
2. what happened;
3. what each party knew;
4. whether representations were made;
5. whether there was breach;
6. whether damage resulted.

The human narrative therefore forms the organisational structure of the dispute.

3. Meaning of Dataset Justice

Dataset justice involves a different evidentiary environment.

Instead of relying predominantly on:

“What did the witness say happened?”

the court may increasingly examine:

“What do the underlying data records demonstrate happened?”

Examples include:

• banking transaction data;
• blockchain transaction histories;
• GPS records;
• access logs;
• smart-contract execution records;
• emails;
• metadata;
• CCTV;
• server logs;
• electronic signatures;
• cloud records;
• automated trading records;
• platform data;
• AI-generated evidence;
• large collections of electronically stored information.

The dataset may contain thousands or millions of individual records.

The challenge therefore shifts from merely collecting evidence to:

selecting, authenticating, contextualising and legally interpreting massive quantities of evidence.

4. Narrative Justice vs Dataset Justice

Importantly, dataset justice should supplement rather than eliminate narrative justice.

A dataset may show what happened technically without explaining:

• why it happened;
• whether it was authorised;
• whether consent existed;
• whether the person understood it;
• whether the data itself is reliable.

5. UAE Legal Context

The UAE is particularly relevant to this transformation because its courts have progressively adopted digital litigation infrastructure.

The DIFC Courts established the Digital Economy Court in 2021 for sophisticated disputes involving technologies such as:

• big data;
• blockchain;
• AI;
• cloud services;
• fintech;
• digital assets;
• robotics;
• automatic dispute-resolution processes. 

The current Digital Economy Court Rules specifically identify complex databases, AI, digitally stored data, blockchain, digital assets and automatic dispute resolution among matters suitable for the specialist court.

Thus, the UAE provides a particularly strong environment for studying the movement from traditional narrative adjudication toward data-intensive adjudication.

6. Dataset Justice Does Not Mean Algorithmic Justice

This distinction is fundamental.

Dataset justice

Means:

Evidence and judicial administration increasingly depend on structured digital information.

Algorithmic justice

Would mean:

An algorithm makes or substantially determines the judicial outcome.

The first is already developing.

The second raises much more serious questions concerning:

• judicial independence;
• procedural fairness;
• explainability;
• accountability;
• bias;
• right to be heard.

The DIFC Courts' own guidance on generative AI makes clear that AI should assist rather than replace the human decision-making integral to litigation. It also requires verification of AI-generated material and warns against over-reliance on such systems.

7. Case Law 1 — DFSA v Commissioner of Data Protection & Waterhouse

Case

Dubai Financial Services Authority v Commissioner of Data Protection & Anna Waterhouse [2018] DIFC CFI 051 and CFI 085

Facts

The dispute concerned a subject-access request for personal information held by the DFSA.

A significant question was whether extensive paper records could fall within the legal concept of a structured filing system and therefore constitute relevant data.

Principle

The DIFC Court examined the legal significance of information organised according to identifiable persons and criteria. The judgment discusses the concept of a filing system in which information relating to an identifiable person is structured so that specific information can be readily accessed.

Significance for Dataset Justice

This case illustrates a foundational principle:

The legal significance of information depends not merely on its physical form but also on how information is organised and made retrievable.

This is central to dataset justice because datasets are essentially structured systems for storing and retrieving information.

8. Case Law 2 — Gate Mena v Tabarak Investment Capital

Case

Gate Mena DMCC (formerly Huobi OTC DMCC) & Huobi Mena FZE v Tabarak Investment Capital Ltd [2024] DIFC DEC 002

Facts

The dispute concerned cryptocurrency trading and the legal consequences of transactions involving crypto-assets.

The Digital Economy Court's 2026 retrial involved extensive expert material and submissions concerning cryptocurrency technology and the treatment of crypto-assets. The parties also produced a compendium of UAE decisions concerning crypto-assets.

Principle

The case demonstrates the necessity of understanding the technical architecture underlying digital transactions before applying conventional civil-law concepts.

Significance

A traditional narrative might be:

“Party A transferred cryptocurrency to Party B.”

A dataset-oriented analysis asks:

• Which wallet initiated the transaction?
• What address received the asset?
• What was the transaction hash?
• When was it recorded?
• What blockchain confirms it?
• Was the transaction authorised?
• What subsequent transfers occurred?

The dispute therefore moves from a purely human narrative to technical transaction reconstruction.

9. Case Law 3 — Gate Mena v Tabarak: Earlier Court of Appeal Proceedings

The same dispute had previously reached the DIFC Court of Appeal in:

Gate Mena DMCC & Huobi Mena FZE v Tabarak Investment Capital Ltd & Christian Thurner [2023] DIFC CA 002.

The Court described cryptocurrency litigation as presenting challenges arising from fraud involving emerging technology and noted the difficulty of allocating losses where the perpetrators of the fraud had disappeared.

Significance

The case demonstrates why dataset-based evidence becomes increasingly important.

In a blockchain dispute, the court may need to reconstruct:

wallet → transaction → intermediary → exchange → subsequent wallet → ultimate destination.

This is much closer to data reconstruction than traditional witness-based fact-finding.

10. Case Law 4 — Techteryx v Aria Commodities

Case

Techteryx Ltd v Aria Commodities DMCC & Others [2025] DIFC DEC 001

Facts

The claimant asserted beneficial ownership of approximately USD 456 million representing reserves backing the TrueUSD stablecoin.

The Digital Economy Court granted proprietary and worldwide freezing relief and subsequently considered disclosure concerning the movement and traceable proceeds of the funds.

Later orders required information concerning onward dealings, current value and location of funds, and the ultimate beneficiaries of relevant assets.

Significance

This is a powerful example of data-intensive civil justice.

Traditional investigation might involve:

witness → document → explanation.

A digital asset investigation may involve:

blockchain record → transaction → wallet → exchange → bank account → subsequent transaction → traceable proceeds.

The dataset itself becomes a map of the alleged movement of property.

This does not mean that the dataset automatically proves liability. The court must still determine ownership, causation, knowledge, tracing and other legal issues.

11. Case Law 5 — Techteryx: Disclosure to Trace Digital Assets

A subsequent Techteryx proceeding concerned applications for disclosure against entities including IG.

The claimant sought information concerning accounts and documents associated with the defendants and related entities. The application relied on the court's jurisdiction to make disclosure orders, including Norwich Pharmacal and Bankers Trust-type relief.

Significance

This illustrates a crucial feature of dataset justice:

The legal system may need to obtain the dataset before the underlying civil claim can be fully reconstructed.

The procedural sequence therefore becomes:

suspected wrongdoing → identification of data → disclosure order → dataset acquisition → transaction reconstruction → legal determination.

This is substantially different from a purely narrative model.

12. Case Law 6 — Nihan v Nicholas & Niaz

Case

Nihan v Nicholas & Niaz [2024] DIFC CA 012

Facts

The case concerned recognition and enforcement of an arbitral award and the interpretation of public-policy and arbitrability provisions.

Principle

The DIFC Court of Appeal carefully distinguished between questions governed by DIFC law and questions concerning UAE public policy in the context of enforcement.

Significance for Dataset Justice

This case demonstrates an important limitation:

Data cannot determine the applicable legal rule by itself.

Even where a court has extensive structured information, it must still perform legal classification.

For example:

Dataset: transaction occurred.

Legal question: Was the transaction legally enforceable?

The second question requires legal reasoning, not simply data processing.

Thus:

Data establishes facts; law gives those facts legal significance.

13. Case Law 7 — Korek Telecom v Iraq Telecom

Case

Korek Telecom Company LLC & Others v Iraq Telecom Ltd & International Holdings Ltd [2024] DIFC CA 016

Facts

The dispute concerned an arbitration involving allegations of tortious conspiracy, governmental action and allegedly unlawfully obtained investigative material.

The DIFC Court of Appeal considered issues concerning the act-of-state doctrine, UAE public policy and evidentiary material.

Significance

This case demonstrates why data provenance is crucial.

In dataset justice, a court must ask:

1. Where did the dataset come from?
2. Who obtained it?
3. Was it obtained lawfully?
4. Has it been altered?
5. Can its authenticity be established?
6. Is it admissible?
7. What weight should it receive?

Therefore:

More data does not automatically mean better evidence.

14. Case Law 8 — DFSA Data Protection Litigation

The DFSA v Commissioner of Data Protection litigation also illustrates another major limitation of dataset justice: privacy and data rights.

The court had to examine the relationship between access to information and the legal framework governing personal data.

The modern data-intensive court therefore faces two simultaneous requirements:

Requirement 1

Obtain sufficient data to decide the dispute.

Requirement 2

Avoid unlawful or disproportionate collection, disclosure or use of personal data.

This creates a central principle:

Dataset justice must remain rights-based justice.

15. The DIFC AI Guidance and Human Judicial Control

Although not a case, the DIFC Courts' Practical Guidance Note No. 2 of 2023 is highly relevant to this topic.

The guidance requires:

• transparency concerning AI use;
• verification of AI-generated content;
• attention to accuracy and reliability;
• disclosure of AI use at an early stage;
• protection of confidentiality;
• consideration of data-protection obligations;
• avoidance of excessive reliance on AI.

Most importantly, it states that AI should assist parties and should not replace the human decision-making required in preparing evidence and submissions.

This provides an institutional boundary around dataset justice.

16. From Witness Credibility to Data Reliability

Traditional justice asks:

“Is this witness telling the truth?”

Dataset justice asks additional questions:

“Is this dataset reliable?”

“Who created it?”

“How was it collected?”

“Was it altered?”

“What algorithm produced it?”

“Are there missing records?”

“Does the dataset contain bias?”

The evidentiary concept of credibility therefore begins to coexist with data integrity.

17. Data Provenance

Data provenance means the ability to establish the history of data.

For example:

Creation → collection → storage → transfer → processing → analysis → presentation

If any stage is compromised, the evidentiary value may be questioned.

This becomes particularly important with:

• blockchain analytics;
• financial datasets;
• AI outputs;
• surveillance data;
• platform records;
• cloud records.

18. Dataset Justice and AI

AI can assist courts or litigants with:

A. Document review

Millions of documents can be classified and searched.

B. Duplicate detection

Identical or substantially identical documents can be identified.

C. Pattern recognition

AI may detect unusual transaction patterns.

D. Timeline reconstruction

Events can be arranged chronologically.

E. Entity identification

AI can connect names, companies, accounts and transactions.

F. Legal research

Large collections of case law can be searched.

The DIFC Courts have already described digital litigation infrastructure incorporating electronic bundles, automated pagination, evidence management and AI-supported document review.

19. Dataset Justice and Big Data

A traditional civil case may involve:

100 documents.

A modern commercial dispute could involve:

millions of emails + financial records + transaction logs + cloud documents + messaging data.

Manual narrative reconstruction becomes increasingly difficult.

Consequently, litigation can move toward:

Data collection → filtering → classification → correlation → visualisation → human legal interpretation.

The judge remains responsible for the final legal determination.

20. Dataset Justice and Blockchain

Blockchain presents an especially strong example.

A conventional contract dispute might depend on:

• testimony;
• invoices;
• correspondence.

A blockchain dispute may additionally contain:

• immutable transaction records;
• wallet addresses;
• timestamps;
• smart-contract events;
• transaction hashes;
• token movements.

This can produce a highly detailed chronological dataset.

However, blockchain proves a transaction occurred on a network; it does not necessarily prove the legal identity of the person controlling an address or the legal purpose of a transaction.

Thus:

Blockchain record ≠ automatic legal conclusion.

21. Dataset Justice and Digital Assets

The Techteryx litigation demonstrates how civil litigation involving stablecoins can require tracing enormous financial movements through multiple systems.

The legal problem may become:

“Who legally owns these assets?”

while the evidentiary problem becomes:

“Where did the assets move?”

The second question may be solved through datasets.

The first still requires legal reasoning.

22. Dataset Justice and Automated Dispute Resolution

The DIFC Digital Economy Court framework expressly includes automatic dispute resolution processes within its potential subject matter.

This raises a future distinction:

Human adjudication

Judge evaluates evidence and gives judgment.

Assisted adjudication

Technology organises and analyses evidence; judge decides.

Automated dispute resolution

Rules/software automatically determine a result.

The last category creates the greatest legal questions concerning:

• explanation;
• appeal;
• accountability;
• procedural fairness;
• human oversight.

23. Dataset Justice and the Right to Be Heard

A dataset can never be treated as self-proving.

A party must be able to challenge:

• the data;
• its source;
• its accuracy;
• its interpretation;
• the algorithm;
• the methodology;
• the statistical inference.

Otherwise, a litigant may face a conclusion that they cannot meaningfully contest.

This is particularly important where AI or automated analytics are used.

24. Dataset Justice and Explainability

Suppose an AI system identifies a party as likely responsible for fraudulent transactions.

The court must not simply accept:

“The algorithm says so.”

The relevant questions include:

• What data was used?
• What methodology was used?
• What assumptions were made?
• What error rate exists?
• What alternative explanations exist?
• Can the conclusion be reproduced?
• Can the opposing party challenge it?

The DIFC AI guidance's emphasis on verification and reliability is directly relevant to this problem.

25. Dataset Bias

Datasets may contain historical or structural bias.

For example:

• previous judicial decisions may reflect historical practices;
• transaction datasets may underrepresent certain events;
• algorithmic classifications may replicate historical patterns;
• incomplete datasets can create misleading correlations.

Therefore:

Historical data is not automatically neutral data.

A court must distinguish between:

correlation

and

legally proven causation.

26. Dataset Justice and Judicial Independence

One of the greatest dangers is allowing a dataset to become a substitute for judicial reasoning.

The correct model should remain:

Dataset → evidence → judicial evaluation → legal reasoning → judgment

not:

Dataset → algorithm → judgment.

This distinction is consistent with the DIFC Courts' stated approach that AI tools should assist litigation but not replace integral human decision-making.

27. Dataset Justice and UAE Civil Law

The new Civil Transactions Law creates the substantive legal framework within which these technological developments must operate.

The new federal Code entered into force on 1 June 2026 and repealed Federal Law No. 5 of 1985.

Accordingly, dataset evidence may increasingly be used to establish facts relevant to:

• contract formation;
• performance;
• breach;
• harmful acts;
• causation;
• damage;
• ownership;
• possession;
• fraud;
• good faith;
• unjust enrichment;
• digital transactions.

But the legal test remains a legal test.

Technology changes how facts are discovered and demonstrated; it does not automatically change the substantive legal rule.

28. Traditional Narrative and Dataset Justice Can Coexist

It would be inaccurate to say that dataset justice completely replaces narrative justice.

A sophisticated civil proceeding may contain both.

Dataset

“The bank records show 17 transfers.”

Narrative

“The defendant claims those transfers were authorised payments under the supply agreement.”

Legal analysis

“The court must determine whether the transfers were made pursuant to the contract and whether the contractual obligation was discharged.”

Thus:

Dataset + Narrative + Law = Judicial determination

29. Evidentiary Hierarchy in Dataset Justice

A useful analytical structure is:

Level 1 — Raw data

Transaction, log, message or record.

Level 2 — Verified data

Authenticity and integrity established.

Level 3 — Interpreted data

Expert explains its meaning.

Level 4 — Factual finding

Court determines what actually happened.

Level 5 — Legal classification

Court applies the applicable legal rule.

Level 6 — Judgment

Court determines rights and remedies.

This prevents the common error of treating an analytical output as equivalent to a judicial finding.

30. Dataset Justice and Expert Evidence

Technical datasets often require experts.

For example:

• blockchain expert;
• cybersecurity expert;
• forensic accountant;
• AI expert;
• data scientist;
• digital-forensics specialist.

The expert explains the technical material.

But:

The expert does not decide the legal dispute.

The court determines the legal consequences of the established facts.

31. Dataset Justice and Civil Liability

Consider an AI-driven trading system.

A claimant alleges:

“The defendant's AI caused AED 20 million in losses.”

A dataset-based investigation could establish:

1. system instructions;
2. algorithmic decisions;
3. timestamps;
4. trades;
5. market movements;
6. system warnings;
7. human interventions;
8. resulting losses.

But the legal question remains:

Who legally bears responsibility?

Possible questions include:

• Who designed the system?
• Who deployed it?
• Who controlled it?
• Was there negligence?
• Was there contractual limitation?
• Was the loss foreseeable?
• Was there an intervening cause?

Thus, dataset analysis supplies the factual architecture for traditional civil-law reasoning.

32. Dataset Justice and Procedural Efficiency

A major advantage is potentially greater efficiency.

AI-supported litigation can help:

• locate relevant documents;
• detect duplicates;
• identify chronological sequences;
• classify evidence;
• identify missing records;
• trace transactions.

The DIFC Courts have already described digital systems incorporating electronic evidence bundles and AI-supported document review.

However, efficiency cannot override:

• due process;
• equality of arms;
• confidentiality;
• privacy;
• evidentiary reliability.

33. Dataset Justice and Privacy

A dataset may contain:

• personal information;
• financial information;
• communications;
• location data;
• biometric information.

Therefore, litigation creates a tension:

The court needs sufficient information to decide the dispute, but the parties' information rights must also be protected.

The DFSA data-protection litigation illustrates the legal importance of defining what information constitutes legally protected data and how it can be accessed.

34. Dataset Justice and Public Order

Dataset-driven adjudication cannot override mandatory legal principles.

The UAE/Federal Supreme Court has emphasised the significance of public-order issues and the binding character of Federal Supreme Court judgments within the federal judicial system. In Cassation No. 250 of 2020, the Court stated that public-order matters may be considered by the court on its own initiative and described its judgments as final and binding.

Thus:

Data determines facts within its evidentiary role; mandatory law determines legal consequences.

35. Main Challenges

1. Data quality

Incorrect data produces incorrect conclusions.

2. Data completeness

Missing data may distort the factual picture.

3. Algorithmic bias

Automated analysis may reproduce hidden biases.

4. Explainability

Parties need to understand significant analytical conclusions.

5. Privacy

Large-scale evidence collection can expose sensitive information.

6. Cybersecurity

Court datasets may themselves become targets for attacks.

7. Authentication

The court must determine whether data is genuine.

8. Human oversight

Judicial responsibility cannot simply be delegated to software.

36. Future Development in UAE Civil Justice

The likely institutional development is not simply:

human courts → robot courts

but:

paper litigation → electronic litigation → data-intensive litigation → AI-assisted adjudication → increasingly specialised digital courts.

The creation and continuing development of the DIFC Digital Economy Court is evidence of this institutional direction.

The court's jurisdiction expressly covers technologies including AI, big data, blockchain, cloud services and automatic dispute resolution.

37. Case-Law Summary

38. Seven Principles of Dataset Justice

Principle 1 — Data is evidence, not judgment

A dataset does not decide the case.

Principle 2 — Provenance matters

The court must know where the data came from.

Principle 3 — Authentication matters

Digital information must be appropriately verified.

Principle 4 — Context matters

Raw data may be meaningless without explanation.

Principle 5 — Explainability matters

Parties should be able to challenge significant analytical conclusions.

Principle 6 — Human adjudication remains essential

AI should assist rather than replace judicial reasoning.

Principle 7 — Rights constrain data use

Privacy, confidentiality, due process and other legal protections remain applicable.

39. Narrative Justice → Dataset Justice: Conceptual Model

Traditional model

Witness → Story → Documents → Judge → Judgment

Transitional model

Witness + Documents + Electronic Records → Judge → Judgment

Dataset-intensive model

Databases + Metadata + Electronic Records + AI Analysis + Expert Evidence → Human Judge → Legal Reasoning → Judgment

Future hybrid model

Data Collection → Automated Organisation → AI-Assisted Analysis → Human Verification → Judicial Evaluation → Reasoned Judgment

This final model is more consistent with the present UAE trajectory than a theory of completely automated adjudication.

40. Important Legal Qualification

The phrase “dataset justice” is a conceptual/academic expression, not a formally recognised category of UAE civil law.

Likewise, the cases discussed above do not establish a UAE legal doctrine called “dataset justice.”

They instead illustrate different components of the transition:

• structured information;
• electronic evidence;
• digital assets;
• blockchain tracing;
• data protection;
• AI-assisted litigation;
• specialised digital courts;
• human judicial oversight.

Most of the technology-specific cases cited here are DIFC cases, not mainland UAE cases. They should therefore not be presented as binding interpretations of the federal Civil Transactions Law.

Similarly, older UAE authorities must be used cautiously because the former 1985 Civil Transactions Law was repealed when the 2025 Civil Transactions Law took effect on 1 June 2026.

41. Exam-Oriented Answer

The transition from narrative justice to dataset justice in UAE civil law refers to the increasing movement from adjudication based primarily on human narratives, witness testimony and individually examined documents toward adjudication supported by large, structured and electronically generated datasets.

The transformation is particularly visible in the DIFC, where the Digital Economy Court handles disputes involving AI, blockchain, big data, digital assets, cloud services and automated dispute resolution.

Cases such as DFSA v Commissioner of Data Protection, Gate Mena v Tabarak, Techteryx v Aria, Nihan v Nicholas & Niaz, and Korek Telecom v Iraq Telecom demonstrate different aspects of this transformation: structured data, cryptocurrency evidence, asset tracing, legal classification, evidentiary provenance and public-policy limitations.

The transformation does not mean that datasets replace judges. Rather, datasets increasingly help courts discover, organise, verify and reconstruct facts, while human judges remain responsible for evaluating evidence and applying the law.

Short formula:

Narrative → Data → Verification → Analysis → Human Evaluation → Legal Reasoning → Judgment

42. Conclusion

The UAE's civil-justice environment is moving toward a model in which data becomes an increasingly important component of fact-finding.

The transformation can be understood as:

Narrative justice asks: “What happened according to the evidence and human accounts?”

while dataset-intensive justice increasingly asks:

“What does the complete digital record demonstrate, and how can that information be legally interpreted?”

The emerging UAE model is therefore best understood as human judicial reasoning supported by increasingly sophisticated datasets, rather than justice delegated to algorithms.

The DIFC's specialist Digital Economy Court, its rules covering AI, big data, blockchain and automated dispute resolution, its AI guidance, and recent digital-asset litigation such as Techteryx demonstrate that this transition is already becoming institutionally significant.

Revision formula:

NARRATIVE → DATASET → AUTHENTICATION → EXPLANATION → HUMAN REVIEW → LEGAL REASONING → JUSTICE