CIVIL LAW AND UAE: SELF-MODIFYING LEGAL SYSTEMS THEORY

1. Introduction

A self-modifying legal system is a theoretical legal-technical system capable of changing its own rules, contractual parameters, decision-making logic, or operational consequences in response to predefined data, events, algorithms, machine-learning outputs, or external conditions.

In ordinary civil law:

Law → Human interpretation → Decision → Legal consequence

In a self-modifying system:

Legal rule → Algorithm → Data → Automatic adaptation → New operational rule → Legal consequence

Examples could include:

a smart contract automatically changing payment obligations after an external economic event;

an insurance platform automatically modifying risk classifications;

an automated dispute-resolution mechanism changing escalation procedures;

an AI system modifying its internal decision model after receiving new data;

a digital platform automatically changing contractual parameters according to predetermined conditions.

The important UAE legal question is therefore:

Can a technological system modify its own operation without thereby acquiring independent authority to modify the parties' legal rights and obligations?

Under current UAE law, the safer answer is that automation may modify performance or trigger contractual mechanisms, but it does not automatically acquire independent legislative, judicial, or contractual authority.

2. Meaning of Self-Modifying Legal Systems

The theory has three principal forms.

A. Self-modifying contractual systems

The contract contains rules allowing its operation to change when specified events occur.

Example:

“If the reference interest rate changes beyond the agreed threshold, the payment formula shall automatically adjust.”

This is potentially compatible with contractual autonomy if the mechanism is sufficiently certain and legally permissible.

B. Self-modifying dispute-resolution systems

A system may automatically move a dispute through different stages:

notification → cure period → technical review → mediation → arbitration → enforcement

This is particularly relevant to smart contracts and automated commercial platforms.

C. Self-modifying decision systems

An AI or algorithm may alter its own model after receiving new data.

This presents a substantially greater legal problem because the system may effectively alter the basis upon which legal consequences are determined.

3. UAE Legal Position

The UAE legal system does not recognize an algorithm as an autonomous source of law merely because it is technically capable of modifying itself.

The distinction should be maintained between:

The fundamental principle is:

A system may execute a legal rule, but technological self-modification does not by itself create legal authority.

4. Current UAE Civil Transactions Framework

A major development is Federal Decree by Law No. 25 of 2025 Promulgating the Civil Transactions Law, which became effective on 1 June 2026.

The new Civil Transactions Law is particularly relevant to technological legal systems because modern civil transactions increasingly involve:

automated contracting;

continuing contractual relationships;

digital performance;

changing economic circumstances;

pre-contractual information;

contractual adjustment;

technologically mediated performance.

The legislation should therefore be understood as providing the legal framework within which adaptive technology operates, rather than allowing technology to become an independent source of civil law.

5. Relationship Between Human Law and Machine Rules

A self-modifying system may contain two different types of rules.

Type 1: Legal rules

These originate from:

legislation;

regulations;

judicial decisions;

valid contractual agreements;

recognized legal principles.

Type 2: Computational rules

These originate from:

software;

algorithms;

smart contracts;

machine-learning models;

databases;

automated decision systems.

The two must not be confused.

For example:

“The algorithm will suspend payment when X occurs”

is a computational rule.

But:

“Suspension of payment constitutes lawful termination of the contract”

is a legal conclusion.

The second proposition requires legal authority.

Therefore:

Code ≠ Law

and:

Automated execution ≠ independent legal authority.

6. Self-Modification and Contractual Autonomy

UAE civil law recognizes the importance of contractual agreement, subject to mandatory legal rules and public policy.

A contract can therefore establish an adaptive mechanism.

For example:

“The price shall automatically adjust according to the agreed published index.”

This is materially different from:

“The software may freely rewrite the price whenever its machine-learning model determines that the price is inappropriate.”

The first establishes an objective contractual formula.

The second potentially delegates excessive discretionary power to an opaque system.

A legally safer adaptive clause should identify:

the triggering event;

the data source;

the calculation methodology;

the maximum adjustment;

the effective date;

notice requirements;

audit rights;

correction mechanisms;

human review;

dispute-resolution procedures.

7. The Principle of Bounded Self-Modification

The strongest theoretical model for UAE civil law is bounded self-modification.

Under this approach, a system may modify itself only within predetermined legal boundaries.

Model

Human-created legal framework

↓

Permitted computational rules

↓

Defined data inputs

↓

Automatic modification

↓

Human/legal supervision

↓

Judicial or arbitral review where disputed

This preserves technological flexibility without allowing the machine to become the ultimate source of legal authority.

8. Self-Modifying Systems and Mandatory Law

A contractual algorithm cannot override mandatory UAE law.

Suppose software provides:

“If the customer fails to pay, all statutory protections are automatically cancelled.”

The computational instruction cannot by itself eliminate mandatory legal protections.

Similarly, an algorithm cannot legally transform:

an unlawful transaction into a lawful one;

an invalid clause into a valid one;

a prohibited activity into a permitted activity;

a judicial dispute into an automatically resolved legal judgment.

This creates a fundamental limitation:

Self-modification is subordinate to the hierarchy of legal norms.

9. Self-Modifying Systems and Public Policy

Public policy creates another major limitation.

An algorithm might attempt to automatically impose:

excessive penalties;

discriminatory terms;

unlawful restrictions;

automatic forfeiture;

unreasonable liability;

disproportionate remedies.

The fact that the result was generated automatically does not make it legally immune from review.

The court can examine the underlying legal relationship and the consequences produced by the system.

10. Electronic Transactions Law

The UAE's Federal Decree-Law No. 46 of 2021 on Electronic Transactions and Trust Services is highly relevant.

It recognizes the legal significance of electronic transactions and provides a framework for transactions involving electronic systems.

Particularly important is the recognition of transactions formed through automated electronic systems.

This demonstrates an important principle:

Human presence at every computational step is not necessarily required for electronic contractual validity.

However, this should not be misunderstood as granting an automated system independent legal personality.

An automated system can facilitate formation or performance without itself becoming the legal person who owns the rights and obligations.

11. Smart Contracts

Self-modifying legal systems are closely connected with smart contracts.

A conventional contract may say:

“Party A shall pay AED 100,000 if condition X occurs.”

A smart contract may automatically transfer the money when condition X is digitally verified.

A self-modifying smart contract goes further:

“If X, Y and Z occur, the system changes the contractual formula itself.”

The legal risk increases as the system moves from:

execution

to

interpretation

and ultimately to:

creation of new obligations.

The first is easier to accommodate within existing law.

The third raises serious questions concerning consent, certainty, authority, fairness, evidence and judicial review.

12. The Role of Oracles

Many self-modifying systems require an oracle.

An oracle supplies external information to the blockchain or automated system.

Examples include:

exchange rates;

commodity prices;

weather conditions;

delivery status;

market indices;

government data.

The critical question becomes:

Who is legally responsible if the oracle provides incorrect information?

Possible answers may involve:

contractual liability;

negligence;

service-provider liability;

breach of warranty;

allocation of technological risk.

Therefore, self-modification does not eliminate traditional civil liability.

It merely changes the mechanism through which the relevant event is detected.

13. Self-Modifying Systems and Evidence

A self-modifying legal system creates special evidentiary problems.

The court may need to establish:

what the original code said;

what modifications occurred;

who authorized them;

what data triggered the modification;

whether the data was accurate;

whether the algorithm was altered;

whether the modification was predictable;

whether the system generated an audit trail;

whether the result can be reproduced.

Therefore, a legally robust system should preserve:

Version history + audit logs + source data + timestamps + authorization records + model changes

Without these records, proving the contractual or technological history may become difficult.

14. Explainability

A self-modifying AI system may change its internal parameters without producing an easily understandable explanation.

This creates an important civil-law issue.

Suppose an AI insurance system changes a person's risk classification.

The affected party may ask:

Why did my contractual position change?

If the answer is simply:

“The model changed itself,”

that may be insufficient where the modification affects legally protected rights.

Consequently:

Explainability becomes part of legal accountability.

The more serious the legal consequence, the stronger the justification for:

transparency;

auditability;

human review;

reasons;

challenge mechanisms.

15. Human-in-the-Loop Principle

A particularly important model for UAE civil-law systems is:

Human-in-the-loop governance.

This means that automated systems may perform routine or objective tasks, while legally significant decisions remain reviewable by an authorized human decision-maker.

For example:

Automatic

calculating invoice amount;

checking payment deadline;

sending reminder;

recording delivery;

comparing objective index values.

Human review

determining fraud;

deciding whether breach is material;

interpreting an ambiguous clause;

determining whether force majeure applies;

determining whether termination is legally justified;

assessing damages.

This distinction is extremely important.

16. Case Law 1 — Techteryx Ltd v Aria Commodities DMCC & Others [2025] DIFC DEC 001

This is one of the most important modern UAE-region authorities for technology-related civil disputes.

The dispute concerned approximately USD 456 million in reserves associated with the TrueUSD stablecoin.

The DIFC Digital Economy Court granted proprietary and worldwide freezing relief concerning the relevant funds and traceable proceeds.

The case demonstrates that even highly technological assets remain subject to conventional legal remedies such as:

proprietary injunctions;

freezing orders;

disclosure obligations;

tracing;

judicial supervision.

Relevance to self-modifying legal systems

Digital infrastructure does not become legally autonomous merely because it operates through technology.

The legal system retains supervisory authority.

Principle

Digital autonomy does not equal legal autonomy.

17. Case Law 2 — Peter Matthew James Gray v Gibson, Dunn & Crutcher LLP [2016] DIFC CA 012

This DIFC Court of Appeal case concerned the relationship between court jurisdiction and an arbitration agreement.

The case is important because it illustrates the importance of giving legal effect to the parties' agreed dispute-resolution architecture.

The parties had structured their relationship around contractual dispute mechanisms, including arbitration.

Relevance

A self-modifying legal system may contain automated escalation mechanisms.

But those mechanisms remain subject to:

the arbitration agreement;

applicable arbitration law;

court jurisdiction;

mandatory procedural requirements.

Principle

Automated dispute architecture operates within legally recognized dispute-resolution authority.

18. Case Law 3 — DAMAC Park Towers Company Limited v Youssef Issa Ward [2015] DIFC CA 006

The dispute involved contractual payments, termination and restitution.

The Court of Appeal examined the contractual arrangement and emphasized the importance of interpreting contractual language in accordance with its commercial context.

The Court rejected an overly expansive interpretation of contractual restitution that could encourage opportunistic termination.

Relevance to self-modifying contracts

An algorithm may mechanically execute a contractual clause.

But the existence of automated execution does not prevent a court from examining:

the contract;

its purpose;

the parties' obligations;

whether termination was lawful;

the legal consequences of termination.

Principle

Machine execution cannot eliminate judicial interpretation of the underlying legal agreement.

This is particularly important where a smart contract automatically terminates a relationship.

19. Case Law 4 — Larmag Holding B.V. v First Abu Dhabi Bank PJSC & Others [2019] DIFC CFI 054

Larmag involved alleged fraudulent misappropriation of valuable bonds and questions concerning proprietary and restitutionary relief.

The case is useful because it demonstrates that courts can examine the underlying legal ownership and movement of assets even where transactions pass through complex financial systems.

Relevance

In a self-modifying financial system, the fact that:

“the system automatically transferred the asset”

does not necessarily answer:

“who was legally entitled to the asset?”

The court can investigate the underlying rights.

Principle

Automated transfer does not conclusively determine legal ownership.

20. Case Law 5 — Arabyads Holding Limited v Gulrez Alam Marghoob Alam [2025] ADGMCFI 0032

This is a particularly important recent UAE-region authority concerning artificial intelligence.

The ADGM Court considered legal work produced with AI assistance in circumstances where authorities relied upon in pleadings were found to be nonexistent, incorrectly cited or unsupported.

The court imposed significant wasted-cost consequences on the legal representatives.

Importance

The case demonstrates that:

Use of AI does not transfer professional responsibility from humans to the technology.

The lawyer remains responsible for verifying the material submitted to court.

Application to self-modifying legal systems

If an AI system changes its own decision-making process, responsibility cannot simply be assigned to the algorithm.

The system requires:

identifiable responsibility;

verification;

supervision;

auditability;

accountability.

Principle

Technological autonomy does not automatically create legal immunity.

21. Case Law 6 — Alawwal Capital JSC v Rasmala Investment Bank Limited [2023] DIFC CFI 038

This dispute concerned investment-related allegations and issues concerning representations and reliance.

The case illustrates the importance of determining objectively:

what representation was made;

whether it was relied upon;

whether reliance was reasonable;

whether the representation caused the alleged loss.

Relevance to adaptive algorithms

Suppose an automated investment system modifies its recommendation after processing new data.

A later dispute may require examination of:

what the system communicated;

what data was available;

what the user understood;

whether reliance occurred;

whether the system's output caused the loss.

Principle

Automated advice does not eliminate causation, reliance and responsibility questions.

22. Case Law 7 — Basin Supply Corporation v Rouge LLC & Claude Barret [2018] DIFC CFI 057

This case concerned a substantial loan and guarantee dispute and involved issues concerning evidence, expert evidence and contractual obligations.

Its broader relevance lies in the court's assessment of objective evidence and legally enforceable obligations rather than merely accepting technical or evidentiary assertions.

Application

A self-modifying system may produce extensive technical records.

But the court must still determine:

authenticity;

relevance;

contractual significance;

evidential weight;

legal consequence.

Principle

Volume of machine-generated data does not itself establish legal truth.

23. Case Law 8 — Youssef Issa Ward v DAMAC Park Towers Company Limited [2014] DIFC CFI 001

The Court of First Instance originally awarded restitution following its interpretation of the parties' contractual arrangement and termination.

The decision was subsequently overturned on appeal in DAMAC v Ward.

Importance for theory

The case is especially useful because it demonstrates why automated systems should not be treated as final interpreters of contractual meaning.

Different judicial interpretations were possible at different stages of the same dispute.

Principle

Where legal meaning is contestable, computational execution cannot substitute for authoritative legal interpretation.

24. What These Cases Collectively Demonstrate

The cases establish an important conceptual pattern:

25. Self-Modifying Systems and Judicial Authority

A fundamental constitutional/civil-law limitation is that an algorithm cannot become a court merely by being sophisticated.

A machine may:

calculate;

classify;

predict;

detect;

recommend;

trigger.

But a judicial decision involves legally authorized adjudication.

Therefore:

Algorithmic recommendation ≠ judgment

and:

Automated enforcement ≠ judicial adjudication

unless a legally recognized mechanism expressly gives the relevant institution that authority.

26. Self-Modifying Systems and Arbitration

Arbitration provides a useful intermediate model.

Parties may agree that certain technical disputes will be resolved through:

expert determination;

arbitration;

institutional procedures;

emergency relief.

An automated system can help identify when an arbitration clause is activated.

But the arbitral tribunal remains the legal decision-maker.

Under UAE Federal Law No. 6 of 2018 on Arbitration, the tribunal possesses legally defined powers, including powers relating to interim and precautionary measures.

Thus:

Software can trigger arbitration.

It does not thereby become the arbitrator.

27. Self-Modification and Mediation

The UAE's Federal Decree-Law No. 40 of 2023 on Mediation and Conciliation in Civil and Commercial Disputes provides another useful framework.

A digital system could automatically:

identify a potential dispute;

notify the parties;

activate a contractual cooling-off period;

recommend mediation;

transmit documents;

schedule mediation.

But settlement ultimately requires legally recognizable agreement.

Therefore:

Automation may facilitate consensual dispute resolution without replacing consent.

28. Adaptive Contracts

The most legally realistic form of self-modifying system is therefore the adaptive contract.

An adaptive contract can contain predetermined adjustment mechanisms.

Example

A five-year supply agreement could state:

If the officially published commodity index increases or decreases by more than 10%, the contract price shall automatically adjust according to a specified formula, subject to a 15% annual cap.

This is substantially more legally predictable than:

The AI may change the contract whenever market conditions appear abnormal.

The first contains:

objective trigger;

objective formula;

defined limit;

foreseeable consequences.

The second contains potentially uncontrolled machine discretion.

29. The Problem of Machine Discretion

Traditional legal systems recognize certain forms of discretion.

Examples include:

judicial discretion;

contractual discretion;

expert discretion;

administrative discretion.

Machine discretion is different because the system may:

process millions of variables;

generate opaque outputs;

modify its internal parameters;

react to data not anticipated by the contracting parties.

Therefore, the legal system should distinguish:

Bounded discretion

from

Unbounded algorithmic discretion.

The first may be manageable.

The second creates substantial legal uncertainty.

30. Predictability and Legal Certainty

Civil law requires parties to understand the consequences of their transactions.

A self-modifying system creates difficulty if parties cannot predict:

future rules;

future calculations;

future modifications;

future liabilities.

Therefore, legal design should include:

modification thresholds;

maximum changes;

predefined variables;

disclosure obligations;

version controls;

audit trails;

termination rights;

human override.

31. Right to Challenge

An important principle should be:

Every materially consequential automated legal outcome should have a legally meaningful challenge mechanism.

For example:

Automated decision

↓

Notice

↓

Explanation

↓

Human review

↓

Mediation/arbitration

↓

Judicial review

This prevents technological finality from becoming legal finality without lawful authority.

32. The “Code Is Law” Problem

The phrase “code is law” is too broad for UAE civil-law analysis.

A more accurate formula is:

Code can operationalize law, but code does not automatically replace law.

For example, blockchain code may prevent a transaction from being reversed technically.

But legal remedies may nevertheless exist if:

consent was defective;

the transaction was induced by fraud;

the underlying agreement was invalid;

mandatory law was violated;

restitution is legally available.

Technical irreversibility and legal irreversibility are therefore different concepts.

33. Self-Modifying Systems and Restitution

Suppose an automated system transfers AED 500,000 because its algorithm incorrectly identifies a contractual condition as satisfied.

The blockchain may make the transfer technically irreversible.

That does not necessarily mean the recipient has a permanent legal right to retain the money.

The current Civil Transactions Law contains rules dealing with restoration of property acquired without lawful basis and recovery of undue performance.

Consequently:

Technical finality ≠ substantive legal finality.

A court may still determine:

whether the transfer had a legal basis;

whether the recipient was entitled to the money;

whether restitution is available;

what remedy is appropriate.

34. Self-Modifying Systems and Liability

A self-modifying system creates several possible responsibility layers.

Layer 1 — Developer

Possible responsibility for:

defective design;

coding errors;

foreseeable vulnerabilities.

Layer 2 — Operator

Possible responsibility for:

improper deployment;

inadequate monitoring;

failure to update safeguards.

Layer 3 — Data provider

Possible responsibility for:

inaccurate data;

defective feeds;

manipulated information.

Layer 4 — Contracting party

Possible responsibility according to contractual allocation of risk.

Layer 5 — Human decision-maker

Possible responsibility where human review or authorization was required.

Thus:

The existence of autonomous software does not eliminate the need to identify a responsible legal person.

35. Self-Modifying Systems and the Legal Person

A major theoretical issue is whether an AI system should itself be treated as a legal person.

Under present UAE civil-law reasoning, technological autonomy does not by itself establish separate legal personality.

The system is generally better understood as:

a tool;

an automated mechanism;

software;

a technological agent acting within authority granted by humans or legal entities.

The rights and liabilities therefore ordinarily remain attached to legally recognized persons.

36. Self-Modifying Systems and Good Faith

Good faith remains important.

A party should not necessarily be permitted to exploit an automated system merely because:

“the code allowed it.”

For example, if a party deliberately manipulates an oracle so that an automatic payment mechanism releases money, the resulting technical execution should not automatically settle the legal question.

The court can investigate:

conduct;

intention where legally relevant;

contractual obligations;

causation;

fraud;

unjust enrichment;

abuse of rights.

37. Self-Modifying Systems and Abuse of Rights

A technologically sophisticated party might design an algorithm to produce an apparently contractual result that is commercially abusive.

For example:

“If the other party misses a payment by one minute, automatically seize all previously paid sums.”

Even if the software executes correctly, the legal enforceability of the resulting consequence may remain subject to applicable law.

Therefore:

Automated contractual enforcement must remain within the boundaries of lawful contractual remedies.

38. The Regulatory Sandbox Model

The UAE's broader technology-regulation environment also supports a controlled experimentation model.

Self-modifying legal systems are better introduced through:

sandbox testing;

limited deployment;

regulatory monitoring;

audit requirements;

human override;

incident reporting;

controlled contractual environments.

A system should not immediately receive unrestricted authority merely because it performs successfully in testing.

39. Recommended Legal Architecture

A UAE-compliant self-modifying system could be designed as follows:

Stage 1 — Legal foundation

Identify:

governing law;

jurisdiction;

mandatory rules;

public-policy limitations.

Stage 2 — Contractual rule

Define:

rights;

duties;

triggers;

adjustment formulas.

Stage 3 — Technical implementation

Translate only sufficiently precise contractual rules into code.

Stage 4 — Data governance

Identify:

data source;

authentication;

reliability;

correction process.

Stage 5 — Controlled modification

Permit changes only within predetermined limits.

Stage 6 — Human review

Require human authorization for legally significant decisions.

Stage 7 — Dispute mechanism

Provide:

negotiation;

mediation;

expert determination;

arbitration;

court access.

Stage 8 — Audit

Preserve:

code versions;

data;

logs;

changes;

approvals.

40. Legal Safeguards

A self-modifying system should ideally contain the following safeguards:

Human override

Immutable audit trail

Version control

Defined modification limits

Source-data verification

Explanation mechanism

Error correction

Notice requirements

Dispute escalation

Emergency suspension

Data-security controls

Responsibility allocation

41. Core Legal Formula

The emerging UAE model can be summarized as:

Human legal authority → contractual authorization → bounded algorithmic execution → monitored modification → human oversight → legal review

Rather than:

AI → autonomous legal rule → automatic finality

The first model is more compatible with existing civil-law principles.

42. Major Legal Problems

Self-modifying legal systems may create the following issues:

1. Consent

Did the parties actually consent to future machine-generated modifications?

2. Certainty

Can the future obligation be identified with sufficient precision?

3. Authority

Who authorized the system to modify the transaction?

4. Accountability

Who is liable when the system makes a harmful change?

5. Evidence

Can the system's historical operation be proved?

6. Explainability

Can affected parties understand the reason for the change?

7. Public policy

Does the automated outcome violate mandatory law?

8. Good faith

Was the system manipulated or exploited?

9. Remedies

Can an automated transaction be reversed legally even if it cannot be reversed technically?

10. Judicial review

Can an affected party challenge the automated result before a court?

43. Advantages

Potential advantages include:

faster contractual adjustment;

reduced transaction costs;

automatic compliance;

continuous monitoring;

reduced administrative errors;

real-time risk management;

automated payment adjustment;

early dispute detection;

improved auditability;

efficient commercial administration.

44. Risks

Major risks include:

algorithmic errors;

unpredictable modification;

data manipulation;

cyberattacks;

model drift;

opaque decision-making;

incorrect oracle information;

unlawful automated termination;

excessive penalties;

loss of human accountability;

conflict between code and law.

45. Distinction Between Self-Executing and Self-Modifying Systems

These concepts should not be confused.

Therefore:

Self-execution is generally easier to reconcile with civil law than unrestricted self-modification.

46. Future UAE Civil-Law Implications

Self-modifying systems may eventually require development in several areas.

Contract law

Rules concerning:

adaptive clauses;

automated modification;

digital consent;

algorithmic interpretation.

Evidence law

Rules concerning:

algorithmic logs;

model outputs;

training data;

reproducibility.

Liability law

Rules concerning:

developer liability;

operator liability;

data-provider liability;

autonomous-system failures.

Procedural law

Rules concerning:

automated dispute resolution;

algorithmic evidence;

remote hearings;

AI-assisted case management.

Arbitration

Rules concerning:

smart-contract arbitration;

automated evidence;

algorithmic expert systems;

emergency technological relief.

47. Critical Principle

The most important legal distinction is:

A system can be autonomous operationally without being autonomous legally.

An AI may operate without human intervention.

A blockchain may execute without human intervention.

A smart contract may modify variables automatically.

But the legal system can still ask:

Who owns the asset?

Was there valid consent?

Was the contract valid?

Was the modification authorized?

Was there breach?

Is the remedy lawful?

Who is responsible?

48. Examination-Oriented Answer

If asked in an examination:

“What is the legal position of self-modifying legal systems in UAE?”

A strong answer would be:

Self-modifying legal systems are not currently recognized in UAE civil law as independent sources of legal authority. They are better understood as technological mechanisms operating within contracts and existing legal frameworks. UAE law permits extensive electronic and automated transactions, but technological automation does not by itself create legal personality or authority to rewrite mandatory law. Adaptive contractual mechanisms may be enforceable where the parties have clearly defined the triggering events, methodology and limits. However, legal interpretation, disputed rights, public policy, mandatory rules and judicial authority remain subject to human legal institutions. Recent DIFC and ADGM decisions concerning digital assets, automated contractual mechanisms and AI-assisted legal work demonstrate that courts continue to apply ordinary principles of contractual interpretation, evidence, accountability and judicial supervision to technologically advanced systems. The preferred model is therefore bounded self-modification with auditability, human oversight and judicial review.

49. Short Revision Notes

Meaning

Self-modifying legal system = system capable of changing its operational rules or legal consequences based on data or algorithmic processes.

Key principle

Automation does not equal legal autonomy.

Main UAE framework

Civil Transactions Law — Federal Decree by Law No. 25 of 2025

Electronic Transactions and Trust Services Law — Federal Decree-Law No. 46 of 2021

Federal Arbitration Law No. 6 of 2018

Mediation and Conciliation Federal Decree-Law No. 40 of 2023

Evidence Law No. 35 of 2022

Civil Procedure Code — Federal Decree-Law No. 42 of 2022

Essential safeguards

Human oversight

Audit trail

Explainability

Version control

Data verification

Defined limits

Challenge mechanism

Judicial review

Key cases

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

Peter Matthew James Gray v Gibson, Dunn & Crutcher LLP [2016] DIFC CA 012

DAMAC Park Towers Co Ltd v Youssef Issa Ward [2015] DIFC CA 006

Larmag Holding B.V. v First Abu Dhabi Bank PJSC & Others [2019] DIFC CFI 054

Arabyads Holding Ltd v Gulrez Alam Marghoob Alam [2025] ADGMCFI 0032

Alawwal Capital JSC v Rasmala Investment Bank Ltd [2023] DIFC CFI 038

Basin Supply Corporation v Rouge LLC & Claude Barret [2018] DIFC CFI 057

Youssef Issa Ward v DAMAC Park Towers Company Ltd [2014] DIFC CFI 001

50. Conclusion

The theory of self-modifying legal systems presents one of the most difficult questions for technologically advanced civil law.

The central UAE position can be expressed in one sentence:

A machine may automatically execute or adapt a legally authorized mechanism, but it cannot acquire independent authority to create, alter or conclusively determine legal rights merely through self-modification.

The future development of UAE civil law is therefore likely to depend not upon choosing between law and technology, but upon establishing a controlled relationship between them:

Law sets the boundaries → contract authorizes the mechanism → technology executes → humans supervise → courts retain ultimate legal authority.

This produces a model of bounded, auditable and reviewable legal automation, rather than unrestricted algorithmic autonomy.