Civil Law And UAE: Epistemology Of Legal Truth In Civil Adjudication

1. Introduction

The epistemology of legal truth in civil adjudication concerns the way a UAE court determines what can be treated as sufficiently established for purposes of a civil judgment.

“Epistemology” is the study of knowledge: how knowledge is obtained, tested, justified and corrected. Applied to civil law, it raises questions such as:

What counts as proof?

How does a court determine whether a fact is established?

How should conflicting evidence be evaluated?

What role should experts play?

How should electronic evidence be authenticated?

Can an AI-generated conclusion be treated as reliable?

How should the court deal with uncertainty?

How does appellate review correct factual or legal error?

Legal truth is therefore not necessarily identical to philosophical or absolute truth.

A civil court normally works with legally admissible and sufficiently persuasive evidence, applies the applicable burden and standard of proof, and reaches a reasoned legal conclusion.

The basic structure can be represented as:

Fact → Evidence → Verification → Judicial evaluation → Legal characterization → Judgment

The reliability of legal truth depends upon every stage of this process.

2. Meaning of Legal Truth

Legal truth can be understood in three different ways.

A. Historical truth

What actually happened in the real world.

B. Evidentiary truth

What can be established from the evidence legally available to the court.

C. Judicial truth

The factual and legal conclusions formally adopted by the court in its judgment.

These three forms of truth may not always coincide.

For example, a transaction may actually have occurred, but if the available evidence cannot establish its terms, the court may be unable to treat those terms as proven.

Therefore:

Legal adjudication does not simply discover reality; it determines legally relevant facts through legally recognized methods of proof.

3. UAE Civil-Law Context

UAE civil adjudication combines several important characteristics:

written legal rules;

documentary evidence;

judicial evaluation of evidence;

court-appointed experts;

witness evidence;

electronic evidence;

contractual interpretation;

statutory interpretation;

procedural safeguards;

appellate review.

The Federal Civil Transactions framework provides the substantive foundation for civil rights and obligations, while procedural legislation determines how disputes are brought and adjudicated.

The DIFC and ADGM courts provide particularly useful modern authorities because their judgments frequently address:

complex commercial evidence;

electronic disclosure;

expert evidence;

arbitration;

digital assets;

procedural fairness;

judicial reasoning.

Those authorities should, however, be distinguished from mainland UAE Supreme Court jurisprudence because the DIFC and ADGM have their own procedural and substantive frameworks in many matters.

4. Legal Truth Is Constructed Through Evidence

A civil court does not ordinarily have direct access to past events.

A dispute may concern something that occurred:

five years ago;

through an electronic system;

between several companies;

through intermediaries;

across multiple jurisdictions.

The court reconstructs the event from available evidence.

For example:

Actual transaction

↓

Documents

↓

Emails

↓

Witness evidence

↓

Expert analysis

↓

Judicial evaluation

↓

Finding of fact

The judicial finding is therefore a reasoned reconstruction of the legally relevant event.

5. Facts and Law Must Be Distinguished

An important feature of civil adjudication is the distinction between:

Questions of fact

Examples:

Was the payment made?

Was the building defective?

Was the email sent?

Was the contract signed?

Did the defendant receive the goods?

Questions of law

Examples:

What does the contractual provision mean?

Was there a breach?

Is the defendant legally liable?

What remedy is available?

Experts can assist with factual and technical matters, but the ultimate legal characterization belongs to the court.

This distinction becomes increasingly important where AI is used.

An AI system might conclude:

“The probability of breach is 85%.”

That is not automatically a legal judgment.

The court must still determine what contractual obligation existed and whether the established facts constitute breach under applicable law.

6. Sources of Legal Truth

The UAE civil adjudication process may draw knowledge from several sources.

6.1 Documentary Evidence

Contracts, invoices, correspondence, bank records and official documents may establish important facts.

6.2 Witness Evidence

Witnesses may explain events that are not fully documented.

6.3 Expert Evidence

Experts assist where specialized knowledge is required.

6.4 Electronic Evidence

Emails, electronic communications, databases, electronic signatures and digital transaction records may establish facts.

6.5 Judicial Inference

The court may draw reasonable conclusions from established circumstances.

6.6 Legal Presumptions

The law may sometimes attach particular consequences to established facts.

7. Case Law

Case 1: Oheo Bank v Parker [2025] DIFC CA 006

This DIFC Court of Appeal authority is important for understanding the relationship between legal truth and procedural fairness.

The dispute involved challenges to an arbitral award, including whether a party had received a reasonable opportunity to present its case and whether the tribunal's reasons were adequate.

The Court emphasized the importance of:

due process;

meaningful opportunity to present a case;

adequate reasons;

practical justice; and

effective appellate review.

Epistemological significance

A conclusion cannot be regarded as procedurally reliable merely because the decision-maker reached it.

The parties must have a meaningful opportunity to place relevant information before the decision-maker and challenge the opposing case.

Therefore:

Procedural fairness is part of the process through which legal truth is established.

This principle has particular importance for AI-assisted adjudication. If an automated system relies upon information that one party cannot meaningfully challenge, the reliability of the resulting conclusion becomes problematic.

8. Case 2: Khaled Salem Musabeh Humad Al Mheiri v John Cameron [2025] DIFC CA 008

The DIFC Court of Appeal addressed issues concerning factual findings, hearsay evidence, legal principles and the reasoning process used by the first-instance court.

The appellate court required adequate explanation of the connection between:

the evidence;

the facts found;

the applicable UAE-law principles; and

the resulting legal conclusions.

Epistemological significance

This case illustrates that judicial truth requires a reasoning chain.

The court should not merely state:

“Fact X is established.”

The judgment should make it possible to understand why Fact X was established.

The epistemological structure is:

Evidence → credibility/reliability assessment → factual finding → legal rule → conclusion

If one of these links is missing, meaningful appellate review becomes more difficult.

9. Case 3: Arabyads Holding Limited v Gulrez Alam Marghoob Alam [2025] ADGMCFI 0032

This ADGM case is particularly significant for modern legal epistemology.

Legal submissions contained authorities that were apparently fictitious or incorrectly cited and apparently resulted from AI-assisted research.

The Court imposed substantial wasted costs on the lawyers involved.

Epistemological significance

The case demonstrates that the appearance of knowledge is not the same as verified knowledge.

An AI system can produce:

a case name;

a citation;

a quotation;

a legal proposition.

But each may require independent verification.

Thus:

AI-generated legal information becomes legally useful only after appropriate human verification.

The case is a powerful modern illustration of the difference between information generation and legally reliable knowledge.

10. Case 4: Taaleem PJSC v National Bonds Corporation PJSC and Deyaar Development PJSC [2010] DIFC CFI 014

This DIFC disclosure decision provides an important example of how courts assess the reliability of searches for documentary truth.

The Court required meaningful information concerning the searches conducted for relevant documents rather than accepting an unsupported assertion that documents had not been found.

Epistemological significance

There is an important distinction between:

“No relevant document exists.”

and:

“No relevant document was found through the search conducted.”

The second proposition does not necessarily prove the first.

This becomes particularly important with:

cloud storage;

email archives;

mobile devices;

WhatsApp;

databases;

deleted files;

backup systems.

Therefore, the method used to discover evidence can itself become an object of judicial scrutiny.

11. Case 5: Anoop Kumar Lal & Paul Patrick Hennessy v Donna Benton [2021] DIFC CFI 005

The case involved further disclosure and searches concerning electronically stored information, including email attachments, private email accounts and WhatsApp communications.

The Court required further investigation and explanation concerning the searches undertaken.

Epistemological significance

This case demonstrates that documentary truth depends not merely on the documents actually produced but also on the quality of the search process.

If the search is defective, the evidentiary picture may be incomplete.

Thus:

Incomplete search → incomplete evidence → potentially incomplete factual reconstruction.

This principle has obvious significance for AI-assisted document review.

12. Case 6: International Electro-Mechanical Services Co. LLC v Emirates Speciality Hospital FZ-LLC [2020] DIFC CFI 114

This construction dispute involved substantial technical and contractual evidence, including UAE-law expert evidence concerning payment certificates.

The judgment considered the legal significance of certification and the relationship between technical evidence and contractual rights.

Epistemological significance

The case demonstrates that expert or technical evidence does not automatically determine the legal outcome.

For example:

Engineer: “The works have been inspected and certified.”

Court: “What legal consequences follow from that certification?”

The second question is a judicial legal question.

This distinction prevents specialized knowledge from becoming an automatic substitute for judicial reasoning.

13. Case 7: Dubai Court of Cassation, Commercial Judgment No. 767 of 2021

UAE jurisprudence concerning court-appointed experts recognizes that experts assist courts principally with technical and factual matters requiring specialized knowledge.

The expert does not ordinarily determine the ultimate legal issue for the court.

Epistemological significance

This provides a useful foundation for understanding AI-assisted adjudication.

An AI system can perform functions comparable to an advanced analytical assistant:

classification;

comparison;

calculation;

pattern recognition;

anomaly detection;

document organization.

But the legal authority to decide the dispute remains with the legally authorized adjudicator.

Therefore:

Analytical capability does not itself create adjudicative authority.

14. Case 8: GFH Capital Ltd v Haigh [2014] DIFC CFI 020

This commercial dispute illustrates the importance of assessing documentary and factual material in its broader commercial context.

Individual documents cannot always be understood independently from:

contractual provisions;

surrounding correspondence;

commercial conduct;

other evidence;

the chronology of events.

Epistemological significance

Legal truth is frequently contextual.

An isolated document may suggest one conclusion, while the complete evidentiary record may produce another.

This is a significant limitation on purely mechanical or automated approaches to legal fact-finding.

15. Case-Law Table

16. Epistemology of Documentary Truth

Documents appear objective, but they can contain:

errors;

omissions;

altered information;

incomplete versions;

misleading descriptions;

incorrect dates;

inaccurate metadata.

Therefore, the court may need to examine:

Authenticity

Is the document genuine?

Integrity

Has it been altered?

Completeness

Is the complete document available?

Attribution

Who created it?

Context

What circumstances explain it?

Legal significance

What legal proposition can reasonably be derived from it?

This is especially important for digital evidence.

17. Epistemology of Electronic Truth

Electronic records create a distinctive problem.

A computer-generated record may appear highly objective because it contains:

timestamps;

system identifiers;

transaction numbers;

metadata.

But the court may still need to establish:

who controlled the account;

whether the system was functioning correctly;

whether data was altered;

whether the timestamp was generated automatically;

whether records were deleted;

whether the database was complete.

Consequently:

Digital existence does not automatically establish legal truth.

18. Epistemology of Expert Truth

Experts provide specialized knowledge.

For example:

engineers explain structural defects;

accountants explain financial calculations;

medical experts explain technical medical issues;

cybersecurity experts explain system compromise;

valuation experts explain market value.

But expert evidence must be distinguished from legal judgment.

The court should consider:

qualifications;

methodology;

factual assumptions;

underlying data;

consistency with other evidence;

competing explanations;

limitations of the expert opinion.

19. Epistemology of AI-Generated Truth

AI creates a new category:

machine-generated inference.

An AI system may identify a pattern that humans did not notice.

However, its output may depend upon:

training data;

input data;

system architecture;

statistical assumptions;

model limitations;

prompt formulation;

incomplete information.

An AI output can therefore be:

accurate, inaccurate, incomplete or contextually misleading.

The court must distinguish:

“The system generated this result.”

from:

“The result is sufficiently reliable to establish this legally relevant fact.”

These are different propositions.

20. Legal Truth and Probabilistic Reasoning

Modern technology frequently expresses conclusions through probabilities.

For example:

“There is an 87% probability that the transaction originated from Account A.”

But civil adjudication requires the court to apply the relevant legal evidentiary framework.

A numerical probability does not automatically answer:

whether a fact is legally established;

whether a contractual condition was satisfied;

whether causation is established;

whether damages are recoverable.

Therefore:

Probability is evidence about uncertainty; it is not automatically the legal conclusion.

21. Legal Truth and Causation

Causation provides another epistemological difficulty.

Suppose:

Cyberattack → system malfunction → incorrect payment → financial loss

Several causes may exist.

The court must determine:

what actually happened;

which actors contributed;

what duties existed;

whether the conduct legally caused the loss;

whether intervening events affected causation;

what loss is legally attributable to the defendant.

A technical system may identify statistical relationships, but legal causation requires legal analysis.

22. Legal Truth and Contractual Interpretation

Contract disputes frequently demonstrate that factual truth alone does not resolve the dispute.

Suppose the court establishes that:

The parties exchanged a particular email.

That establishes a fact.

The next question may be:

What legal meaning does the email have?

The answer may depend upon:

the contract;

applicable statutory rules;

commercial context;

subsequent conduct;

the parties' obligations.

Thus:

Fact-finding and legal interpretation are separate epistemic stages.

23. Legal Truth and Burden of Proof

A party asserting a civil right generally must establish the factual basis of its claim according to the applicable evidentiary rules.

The burden of proof has an epistemological function.

It determines:

Who must establish the relevant proposition when the evidence remains uncertain?

For example, if two competing explanations remain possible and the applicable law places the burden on the claimant, uncertainty may prevent the claimant from establishing the necessary fact.

Thus, legal truth is affected not only by evidence but also by the legal allocation of evidentiary burdens.

24. Judicial Inference

Courts do not always receive direct evidence.

They may use circumstantial evidence.

For example:

a payment is made immediately before a contractual deadline;

communications show knowledge of the transaction;

company records identify a particular decision-maker;

subsequent conduct is consistent with one explanation.

The court may infer a fact from a collection of circumstances.

But inference should not become speculation.

The distinction is:

Inference = conclusion reasonably supported by established circumstances.

Speculation = conclusion not sufficiently supported by evidence.

25. The Problem of Incomplete Truth

Civil adjudication frequently occurs under conditions of incomplete information.

Documents may be:

destroyed;

unavailable;

confidential;

held by third parties;

stored overseas;

encrypted;

deleted;

inaccessible.

Consequently, the court may need to decide a dispute without possessing every piece of information that existed in reality.

This does not necessarily invalidate the adjudication.

Instead, the legal system uses:

disclosure;

evidentiary rules;

presumptions;

expert evidence;

witness evidence;

procedural orders;

judicial inference;

appellate review.

These mechanisms reduce the effect of informational incompleteness.

26. Judicial Reasons as a Truth-Control Mechanism

Reasoned judgments are central to legal epistemology.

A judgment should ordinarily make it possible to understand:

What facts were established?

↓

What evidence supported them?

↓

Which evidence was rejected?

↓

What law applies?

↓

How does the law apply to the facts?

↓

Why does the remedy follow?

This creates an audit trail of legal reasoning.

Without sufficient reasons, errors may be difficult to detect.

The reasoning requirements emphasized in Oheo Bank v Parker and Al Mheiri v Cameron therefore have broader epistemological importance.

27. Appellate Review and Correction of Legal Truth

First-instance adjudication is not necessarily the final stage of legal reasoning.

Appellate review can identify:

errors of law;

inadequate reasoning;

procedural unfairness;

misappreciation of evidence;

failure to address material arguments;

incorrect legal characterization.

Therefore:

Trial court → appellate review → possible correction

is itself part of the legal system's epistemic architecture.

The purpose is not to guarantee perfect truth but to create institutional mechanisms for correcting significant error.

28. Human and Machine Epistemology

Human and machine reasoning possess different characteristics.

The appropriate legal model is therefore not necessarily:

Human versus machine

but:

Machine assistance + human verification + judicial reasoning.

29. Epistemic Conflict

Conflict may arise when:

Human interpretation ≠ machine prediction

For example:

AI predicts that a contract clause has a high probability of being interpreted in a particular way based on thousands of cases.

The judge may reach a different conclusion because:

the current statutory language changed;

the contract contains unusual wording;

the factual context differs;

a recent binding authority changed the law;

the machine relied on outdated cases.

The machine's statistical result does not automatically override the legally authorized judicial conclusion.

30. Epistemic Reliability of Legal Databases

Legal databases themselves can produce errors.

A database may contain:

outdated judgments;

incomplete case histories;

incorrect headnotes;

superseded legislation;

cases subsequently distinguished or overruled.

Therefore, legal researchers should verify important propositions against authoritative legal materials.

The lesson from Arabyads v Alam is particularly relevant:

The existence of apparently authoritative legal information is not equivalent to verification of that information.

31. UAE Civil Adjudication and the New Civil Transactions Law

The UAE's new Federal Decree by Law No. 25 of 2025 Promulgating the Civil Transactions Law, effective from 1 June 2026, makes temporal accuracy particularly important in contemporary civil litigation.

A legal researcher or AI system must distinguish between:

the 1985 Civil Transactions Law;

the 2025 Civil Transactions Law;

transitional provisions;

judgments interpreting the earlier legislation;

judgments applying the new legislation.

Consequently, an apparently correct historical proposition may not automatically represent the current legal position.

This demonstrates that temporal accuracy is itself part of legal truth.

32. Practical Example: Digital Contract Dispute

Assume a UAE company alleges that a supplier accepted a contract electronically.

The claimant produces:

an email;

an electronic signature;

server logs;

an AI report stating that the defendant's account generated the signature.

A reliable adjudication would proceed as follows:

Stage 1 — Authentication

Is the electronic signature attributable to the defendant?

Stage 2 — Technical evidence

Are the server records reliable?

Stage 3 — AI analysis

How did the AI reach its conclusion?

Stage 4 — Challenge

Can the defendant dispute the methodology?

Stage 5 — Contractual analysis

What agreement was actually formed?

Stage 6 — Legal characterization

Was there a legally binding obligation?

Stage 7 — Remedy

What loss resulted from any breach?

Thus, the AI report is one component of the evidentiary chain, not the complete legal truth.

33. Practical Example: Construction Dispute

Suppose an engineer concludes that a building defect resulted from defective construction.

The court should distinguish:

Technical proposition:

“There is a structural defect.”

from:

Legal proposition:

“The defendant is legally responsible for the defect.”

The second proposition may depend upon:

contractual scope;

engineer's role;

contractor's obligations;

applicable construction liability provisions;

causation;

statutory liability;

limitation periods.

This is why expert evidence cannot simply substitute for legal adjudication.

34. Epistemic Safeguards in UAE Civil Adjudication

A reliable civil adjudication system can use several safeguards:

1. Disclosure

Relevant documents should be identified and produced according to applicable procedural rules.

2. Expert scrutiny

Technical evidence can be tested.

3. Contradictory submissions

Opposing parties can challenge evidence.

4. Judicial reasoning

Material conclusions should be explained.

5. Appellate review

Errors can be corrected.

6. Human verification of AI

Machine-generated authorities and analysis should be checked.

7. Temporal legal verification

Current legislation must be distinguished from repealed or superseded legislation.

35. Epistemic Failure in Civil Adjudication

Epistemic failure can occur where:

evidence is fabricated;

evidence is incomplete;

electronic searches are inadequate;

an expert uses defective methodology;

AI invents authorities;

outdated law is applied;

relevant evidence is ignored;

witness evidence is misunderstood;

technical evidence is treated as legal judgment;

reasons fail to explain material conclusions.

The existence of an epistemic error does not automatically mean the judgment is legally invalid. The legal consequence depends upon the type and seriousness of the error and the applicable procedural rules.

36. Future UAE Civil Justice

Future civil adjudication may increasingly involve:

AI-assisted research;

predictive analytics;

automated document review;

blockchain evidence;

digital identity evidence;

algorithmic financial analysis;

automated valuation;

electronic expert systems;

AI-assisted arbitration;

machine-generated chronologies.

The fundamental legal challenge will be:

How can technological information be integrated into adjudication without confusing computational output with legally established truth?

A sound approach is:

Machine-generated information

↓

Authentication

↓

Human verification

↓

Opposing-party challenge

↓

Judicial evaluation

↓

Reasoned legal conclusion

37. Core Principles

The epistemology of legal truth in UAE civil adjudication can be summarized through twelve principles:

Legal truth is established through legally recognized processes of proof.

Historical truth and judicially established truth are not necessarily identical.

Evidence must be evaluated for authenticity, relevance and reliability.

Electronic evidence requires appropriate verification.

Expert evidence assists the court but does not replace legal judgment.

Circumstantial evidence may support reasonable inference but not speculation.

AI-generated information requires appropriate human verification.

Machine probability does not automatically constitute legal proof.

Parties must have appropriate opportunities to challenge material evidence.

Judicial reasons provide an important mechanism for testing factual and legal conclusions.

Appellate review provides an institutional mechanism for correcting legal error.

The applicable law must be identified accurately in time as well as substance.

38. Conclusion

The epistemology of legal truth in UAE civil adjudication concerns the methods by which courts transform disputed information into legally recognized findings of fact and ultimately into enforceable legal judgments.

The central distinction is:

What actually happened is a question of reality; what can legally be established is a question of evidence; and what legal consequences follow is a question of judicial reasoning.

The UAE and DIFC/ADGM authorities demonstrate this through different dimensions.

Oheo Bank v Parker emphasizes procedural fairness and the importance of adequate reasoning.

Al Mheiri v Cameron demonstrates the need for an identifiable connection between evidence, factual findings and legal reasoning.

Arabyads v Alam demonstrates that AI-generated legal information must be independently verified.

Taaleem and Lal & Hennessy demonstrate that the reliability of electronic truth depends partly on the adequacy of the search process.

International Electro-Mechanical Services and the UAE expert-evidence jurisprudence demonstrate the distinction between technical knowledge and ultimate legal judgment.

The resulting model is:

Evidence → authentication → testing → inference → judicial reasoning → legal characterization → judgment → review.

AI, electronic evidence and advanced analytical systems can improve the ability of courts and litigants to process information. They do not, however, eliminate the need for human legal judgment.

Ultimately, the epistemology of UAE civil adjudication rests upon a fundamental principle:

Legal truth is not simply whatever information is available; it is the conclusion that the legally authorized decision-maker reaches through a procedurally fair, evidentially justified and legally reasoned process.