Civil Law and UAE: Self-Learning Legal Systems and Doctrinal Drift

1. Meaning of Self-Learning Legal Systems

A self-learning legal system is a legal system capable of adapting its interpretation and application through:

judicial decisions;

appellate review;

reconsideration of earlier interpretations;

legislative amendments;

development of procedural principles;

treatment of new technologies and commercial practices;

correction of judicial errors; and

accumulation of precedents and institutional experience.

It does not necessarily mean that artificial intelligence independently makes law. In UAE civil law, the more useful concept is an institutionally learning legal system: courts, legislators, regulators and legal institutions learn from previous disputes and progressively refine the application of legal rules.

This is particularly relevant in the UAE because the legal environment combines codified federal law, emirate-level legislation, specialised jurisdictions such as the DIFC and ADGM, and rapidly developing commercial and technological sectors.

2. Meaning of Doctrinal Drift

Doctrinal drift occurs when the practical meaning or application of a legal doctrine gradually changes over time, even though the statutory language may remain substantially unchanged.

For example:

Statutory rule → early interpretation → repeated judicial application → new factual circumstances → modified interpretation → new doctrine.

Drift can be:

Legitimate doctrinal development

Courts adapt established principles to new circumstances while remaining within statutory authority.

Uncontrolled doctrinal drift

Judicial interpretations move so far from statutory language or legislative purpose that predictability and legal certainty become difficult.

Therefore, the central problem is balancing:

Adaptability + consistency + legality + certainty.

3. UAE Civil-Law Framework

The UAE is primarily a codified civil-law jurisdiction. This means that legislation remains the principal source of legal rules.

The new Federal Decree by Law No. 25 of 2025 Promulgating the Civil Transactions Law, effective from 1 June 2026, replaced the former 1985 Civil Transactions Law. This itself demonstrates institutional legal learning: the legislature can revise the framework when accumulated experience, commercial developments and social conditions justify reform.

The new Code's Article 120 is particularly relevant to doctrinal development. It directs courts to consider contractual consent and commitments, intentions and meanings, literal wording, commercial customs, justice and good faith, and the circumstances surrounding the contract. (UAE Legislation)

Thus, codification does not mean that every legal question is mechanically solved by reading one statutory sentence.

4. How a UAE Legal System Can "Learn"

A simplified model is:

Legislation
↓
Judicial interpretation
↓
Application to real disputes
↓
Appeal/review
↓
Identification of error or uncertainty
↓
Doctrinal clarification
↓
Legislative response where necessary
↓
New legal baseline

This is a form of institutional learning.

It is especially visible in specialised UAE jurisdictions such as the DIFC, where published appellate judgments expressly consider precedent, previous decisions and the limits of judicial law development.

5. Doctrinal Drift in a Codified System

Doctrinal drift can occur even without a common-law doctrine of binding precedent.

It may arise through:

repeated interpretation of statutory provisions;

judicial treatment of commercial customs;

changing understanding of good faith;

new technological circumstances;

changing procedural expectations;

interaction between federal and specialised jurisdictions;

legislative amendments;

appellate correction of first-instance decisions.

For example, the current Civil Transactions Law expressly incorporates commercial custom, good faith, justice and surrounding circumstances into contractual interpretation. These concepts necessarily require judicial application to changing factual environments. (UAE Legislation)

6. Difference Between Self-Learning and Doctrinal Drift

The important point is that self-learning is a process, while doctrinal drift is a possible consequence of that process.

7. Important UAE/DIFC Case Laws

Because there are relatively few reported mainland UAE cases explicitly discussing the theoretical expression "self-learning legal systems", the most useful authorities come from UAE specialised courts, particularly the DIFC Courts. These cases should not be treated as automatically binding mainland UAE precedents. They are valuable for understanding how legal doctrine develops, corrects itself and maintains institutional continuity.

Case 1: Carmon Reestrutura-Engenharia v Cuenda [2024] DIFC CA 003

This is one of the most important authorities for the subject.

The DIFC Court of Appeal directly considered the place of precedent and the circumstances in which an appellate court may depart from an earlier decision. The case concerned the correctness of an earlier decision, Sandra Holding Ltd v Al Saleh & Ors.

The Court considered the proper limits of judicial development because the DIFC Courts are statutory courts whose jurisdiction and powers derive from legislation. (DIFC Courts)

Importance

This illustrates controlled self-learning:

Previous doctrine → identification of possible error → appellate reconsideration → correction or confirmation.

The Court recognised that judicial development cannot become unlimited judicial legislation.

Principle

A legal system can learn from its own precedents, but that learning must remain within the authority granted by legislation.

Case 2: Sandra Holding Ltd v Al Saleh & Ors

The Sandra Holding doctrine became the subject of reconsideration in Carmon.

Its significance therefore lies not only in the original decision but in the institutional feedback loop created when a later court asks whether an earlier doctrine should continue to be followed.

This demonstrates an important characteristic of doctrinal drift:

A doctrine can become established through repetition, but repetition does not necessarily make an interpretation permanently immune from reconsideration.

The Carmon Court specifically identified questions concerning whether Sandra Holding had been wrongly decided and examined the limits of judicial development. (DIFC Courts)

Case 3: Lachesis v Lacrosse [2021] DIFC CA 005

In Lachesis v Lacrosse, the DIFC Court of Appeal considered a challenge to an arbitral award.

The Court dealt with the limited statutory grounds for intervention in arbitration and ultimately dismissed the appeal. (DIFC Courts)

Relevance to self-learning

A legal system does not learn by correcting every allegedly incorrect decision.

It must distinguish between:

genuine procedural/legal defects; and

disagreement with the substantive result.

This creates an important error-filtering mechanism.

If every dissatisfied litigant could reopen the merits of a final decision, the system would not become more intelligent—it would become unstable.

Case 4: Oheo Bank v Parker [2025] DIFC CA 006

This is a particularly useful recent example.

The DIFC Court of Appeal considered challenges to a DIAC arbitral award under Article 41 of the DIFC Arbitration Law. The Court described court intervention as subject to a high threshold and considered whether procedural fairness and the scope of the arbitration justified intervention. (DIFC Courts)

The Court ultimately allowed certain grounds of appeal and set aside specified portions of the award while rejecting another ground. (DIFC Courts)

Relevance

This demonstrates selective correction.

A self-correcting system does not necessarily replace an entire decision when one component is defective.

Instead:

valid reasoning → preserved
defective reasoning → corrected
unaffected portions → retained.

This is an important safeguard against excessive doctrinal disruption.

Case 5: LXT Real Estate Broker LLC v SIR Real Estate LLC [2026] DIFC CA 005

The DIFC Court of Appeal dealt with an appeal involving security for costs and reconsideration of the lower court's approach.

The Court's published materials show the appellate process being used to reconsider matters and remit issues for fresh consideration rather than treating every disputed issue as permanently settled. (DIFC Courts)

Relevance

This illustrates another form of legal-system learning:

Appeal → identification of error → remittal → fresh evaluation.

Remittal is particularly important because some questions depend on factual circumstances and should be reconsidered by the appropriate first-instance decision-maker rather than permanently determined at appellate level.

Case 6: Earlene v Earl [2014] DIFC CFI 011

This case illustrates the opposite side of self-learning: finality.

The doctrine of res judicata prevents a court from repeatedly reconsidering matters that have already been finally determined.

Importance

A legal system cannot function as a permanently self-revising machine.

There must be a point at which:

dispute → decision → appeal opportunity → finality.

Otherwise, legal doctrine would continuously drift because every final judgment could be reopened.

Therefore, res judicata acts as a memory mechanism of the legal system.

It preserves institutional learning by preventing the system from forgetting its previous final decisions.

Case 7: BAM Higgs & Hill LLC v Affan Innovative Structures LLC

The BAM Higgs & Hill litigation demonstrates the continuing procedural lifecycle of a dispute within the DIFC Courts, including trial, judgment and subsequent appellate-permission proceedings. The DIFC Court records show a 2026 judgment followed by an application for permission to appeal. (DIFC Courts)

Relevance

This demonstrates that doctrinal development is not confined to the final judgment.

A legal system learns through:

pleadings;

evidence;

procedural rulings;

trial judgments;

appeal applications;

appellate review.

Each stage can clarify how procedural and substantive rules should operate.

8. Role of the New UAE Civil Transactions Law

The new Civil Transactions Law provides an important foundation for controlled doctrinal development.

Article 120 contains interpretive principles dealing with:

contractual consent;

intentions;

literal meaning;

express statements;

commercial customs;

justice;

good faith;

surrounding circumstances;

protection of weaker parties in cases of ambiguity. (UAE Legislation)

This means that legal interpretation necessarily interacts with factual and commercial developments.

Example

Suppose a commercial contract uses a technological term that was uncommon when the contract was drafted.

A rigid approach might focus exclusively on the dictionary meaning.

A contextual approach may examine:

contractual purpose;

commercial practice;

industry custom;

parties' conduct;

surrounding circumstances.

The legal rule has not necessarily changed. Its application has become more sophisticated.

That is a form of institutional learning.

9. Doctrinal Drift Through Good Faith

Good faith is particularly capable of producing doctrinal development.

It may influence:

contractual performance;

interpretation;

exercise of contractual rights;

cooperation obligations;

prevention of opportunistic conduct;

assessment of contractual behaviour.

But there is a danger.

If courts use broad concepts such as good faith without sufficiently connecting them to legislation and established principles, parties may find it difficult to predict outcomes.

Therefore:

Good faith should develop doctrine, not replace legislation.

10. Doctrinal Drift Through Technology

Technology creates an especially strong pressure for legal adaptation.

Examples include:

AI-generated evidence;

automated contracting;

smart contracts;

blockchain records;

autonomous systems;

digital assets;

algorithmic decision-making;

electronic signatures;

online dispute resolution.

The statute may have been drafted without specifically mentioning the technology.

Courts must therefore determine whether existing concepts can accommodate the new facts.

For example:

"Electronic record"
↓
blockchain record
↓
automated transaction
↓
dispute concerning attribution
↓
existing evidentiary and contractual principles applied to new technology.

This is not necessarily creation of a new legal rule. It may be technological adaptation of an existing doctrine.

11. Self-Learning and AI-Assisted Justice

The concept becomes particularly important when AI is introduced into legal institutions.

A genuine self-learning legal system should not mean that an AI system independently changes legal doctrine.

Instead, AI could assist with:

identifying inconsistent judgments;

detecting recurring legal issues;

comparing statutory interpretations;

identifying contradictory reasoning;

finding emerging patterns;

monitoring changes in legislation;

assisting judges with research.

But the final doctrinal decision should remain subject to legally authorised human institutions.

The Carmon decision is particularly instructive because it emphasises that the DIFC Courts are statutory courts and therefore judicial development has legal boundaries. (DIFC Courts)

12. Human Oversight Is Essential

A self-learning legal system requires several safeguards:

1. Human judicial responsibility

AI should not become the final source of legal authority.

2. Transparent reasoning

Parties should understand why a doctrine has been applied or changed.

3. Appeal

Incorrect interpretations must be challengeable.

4. Precedent management

Courts should distinguish between established doctrine and emerging interpretation.

5. Legislative supremacy

Where a doctrinal issue requires major policy change, legislation may be more appropriate than judicial innovation.

6. Finality

Not every decision should remain permanently open to reconsideration.

13. Difference Between Doctrinal Development and Judicial Law-Making

This distinction is crucial.

The UAE context requires particular caution because civil-law systems place significant importance on written legislation.

The Carmon judgment expressly recognised the question of the proper limits of judicial, as distinct from legislative, development of law. (DIFC Courts)

14. Self-Learning Through Appellate Correction

A useful UAE model is:

Stage 1 — Initial rule

A court interprets legislation.

Stage 2 — Repeated application

Other courts apply the reasoning.

Stage 3 — Problem appears

A new case exposes inconsistency or an unforeseen consequence.

Stage 4 — Appeal

The higher court examines the reasoning.

Stage 5 — Correction

The court confirms, modifies, distinguishes or abandons the previous interpretation where legally justified.

Stage 6 — Stabilisation

The new interpretation becomes part of the legal environment.

Stage 7 — Legislative response

If necessary, the legislature clarifies or changes the statutory rule.

This is essentially a legal feedback loop.

15. Role of Finality in Preventing Excessive Drift

Self-learning without finality would be dangerous.

Imagine:

Judgment A → reconsidered → Judgment B → reconsidered → Judgment C → reconsidered → Judgment D

If this continues indefinitely, citizens cannot reliably organise their affairs around legal rules.

Finality therefore performs an important function.

Finality provides:

certainty;

reliance;

procedural economy;

respect for judgments;

stability of commercial transactions.

The principle of res judicata is therefore not opposed to legal learning. It is one of the conditions that makes controlled legal learning possible.

16. Role of Arbitration

Arbitration provides another example of controlled doctrinal evolution.

The DIFC arbitration framework limits judicial intervention. Lachesis demonstrates that courts will not simply reconsider the merits of an arbitral award because a party disagrees with the tribunal. (DIFC Courts)

Similarly, Oheo Bank illustrates that intervention is tied to specified statutory grounds and procedural fairness rather than functioning as an unrestricted appeal on every legal or factual question. (DIFC Courts)

Thus:

Arbitration autonomy + limited judicial review + exceptional correction = controlled legal adaptation.

17. Main Risks of Doctrinal Drift in UAE

A. Predictability risk

Businesses may find it difficult to determine the meaning of a legal rule if interpretations change too rapidly.

B. Fragmentation

Different legal environments—mainland UAE, DIFC and ADGM—may develop different approaches.

C. Technology-driven uncertainty

New technologies can expose gaps between old legislation and new commercial reality.

D. Judicial overextension

Courts may face difficult questions about whether they are interpreting law or effectively creating new rules.

E. Unequal development

Some areas of commercial law may receive extensive judicial development while others remain comparatively underdeveloped.

18. Advantages of Controlled Doctrinal Development

Properly controlled development can produce:

greater legal adaptability;

better treatment of novel commercial disputes;

correction of judicial mistakes;

greater coherence between different cases;

responsiveness to technological change;

improved interpretation of general statutory concepts;

better alignment between legal rules and commercial reality.

The Carmon litigation is particularly important because it demonstrates that an appellate court can confront the question of whether an earlier interpretation remains appropriate rather than treating precedent as completely immutable. (DIFC Courts)

19. Practical UAE Example

Suppose UAE legislation establishes a general rule concerning contractual performance.

Initially:

Court A interprets the rule narrowly.

Later:

Court B encounters a complex AI-generated contractual process.

The old interpretation does not easily address the new circumstances.

The court may examine:

statutory language;

legislative purpose;

contractual intention;

good faith;

commercial custom;

technological context;

existing judicial decisions.

The court may then:

follow → distinguish → clarify → or, where legally permissible, modify the previous interpretation.

That is doctrinal learning.

It becomes problematic only if the new interpretation cannot reasonably be reconciled with the statutory framework.

20. Seven Case Laws — Quick Revision Table

The reported DIFC materials confirm the appellate role in Carmon, Lachesis, Oheo Bank and LXT, while Carmon expressly addresses the relationship between precedent and judicial development. (DIFC Courts)

21. Key Legal Principles

Principle 1

A codified legal system can still develop through interpretation.

Principle 2

Self-learning does not mean unlimited judicial creativity.

Principle 3

Precedent can provide continuity while still permitting correction in appropriate circumstances.

Principle 4

Appeals provide an institutional mechanism for correcting doctrinal errors.

Principle 5

Res judicata prevents excessive doctrinal instability.

Principle 6

Legislative amendment is the strongest formal mechanism for major doctrinal change.

Principle 7

Technology increases pressure for doctrinal adaptation.

Principle 8

AI may assist legal learning, but legal authority should remain attributable to legally authorised institutions.

22. Relationship Between Self-Learning and Doctrinal Drift

The relationship can be represented as:

New social/commercial facts
↓
Existing legal rule
↓
Judicial application
↓
Repeated interpretation
↓
Emergence of doctrine
↓
New facts expose limitations
↓
Appellate reconsideration
↓
Correction / distinction / confirmation
↓
Doctrinal evolution

Therefore, doctrinal drift is not automatically a defect.

It becomes legally problematic when change is:

unpredictable;

inconsistent;

unsupported by statutory authority;

insufficiently reasoned;

incompatible with finality; or

effectively legislative without legislative authority.

23. Conclusion

Self-learning legal systems and doctrinal drift in UAE civil law describe the tension between stability and adaptation.

The UAE's codified framework provides the foundation, while courts develop its application through interpretation, appeals and accumulated experience. The new Civil Transactions Law reinforces contextual interpretation by expressly recognising matters such as intention, commercial custom, good faith and surrounding circumstances. (UAE Legislation)

The DIFC jurisprudence provides particularly clear examples. Carmon demonstrates that precedent itself can be examined and potentially corrected, while recognising the statutory limits of judicial development. Oheo Bank and Lachesis demonstrate that correction must operate within defined procedural boundaries, and res judicata provides the necessary endpoint to the process.

Thus, the ideal UAE model can be summarised as:

Codification + Judicial Learning + Appellate Correction + Legislative Oversight + Finality = Controlled Doctrinal Evolution

Exam-ready formula

Self-Learning Legal System = Adaptation + Precedent + Review + Correction + Legislative Control + Finality

Doctrinal Drift = Gradual Change in Legal Interpretation

Healthy Legal Development = Doctrinal Adaptation Without Loss of Legality or Predictability.