Civil Law and UAE Shift from Adjudication to Automated Resolution

1. Introduction

The shift from adjudication to automated resolution refers to the movement from a traditional model—where a judge, arbitrator, or mediator personally determines a dispute—to a technology-assisted or potentially technology-determined model in which software, algorithms, smart contracts, artificial intelligence (AI), digital platforms and automated decision systems perform some or all stages of dispute resolution.

In the UAE, this shift is developing, but it is important to state the present legal position accurately:

UAE law presently supports digitalisation and technology-assisted dispute resolution, but it has not generally replaced human judicial decision-making with autonomous AI adjudication.

Current UAE practice uses technology for electronic filing, virtual proceedings, document management, evidence processing and administrative functions. The DIFC has gone further by establishing a Digital Economy Court with rules expressly contemplating artificial-intelligence-driven forms and decision-tree systems. Nevertheless, the actual determination of legal disputes remains subject to judicial or arbitral authority. (Global Practice Guides)

2. Meaning of Automated Resolution

Automated resolution can operate at several different levels.

Level 1 — Digital administration

Technology performs:

filing;

scheduling;

notification;

document storage;

case tracking;

payment processing.

Human adjudication remains completely intact.

Level 2 — AI-assisted adjudication

AI assists the decision-maker by:

searching authorities;

summarising documents;

identifying disputed issues;

analysing evidence;

detecting inconsistencies;

calculating contractual amounts.

The judge or arbitrator remains responsible for the decision.

Level 3 — Automated recommendation

An algorithm analyses the dispute and recommends:

settlement amount;

liability outcome;

procedural direction;

likely interpretation.

A human may accept, reject or modify the recommendation.

Level 4 — Automated determination

The software itself determines the contractual or legal outcome.

Level 5 — Self-executing resolution

The system not only determines the result but also executes it automatically.

For example:

Payment default → algorithm verifies default → smart contract transfers digital asset → dispute automatically closes.

The final two levels create the greatest legal difficulties.

3. UAE's Present Position

There is currently no general UAE statute establishing AI as an autonomous replacement for judges or arbitrators.

Current UAE dispute-resolution law instead provides a framework in which technology can support the legal process.

The UAE Evidence Law, Federal Decree-Law No. 35 of 2022, expressly recognises electronic evidence, including electronic instruments, signatures, correspondence, modern communications, electronic media and other electronic evidence. It also provides rules concerning the evidentiary value of electronically generated records. (UAE Legislation)

Similarly, current commentary on UAE dispute resolution reports that there is no dedicated onshore UAE statute specifically regulating AI in litigation, arbitration or ADR; existing evidence, procedure, confidentiality, data-protection and professional obligations therefore remain relevant. (Chambers)

4. Why the UAE Is Moving Towards Automated Resolution

Several factors encourage automation.

1. Speed

Algorithms can process large quantities of documents much faster than humans.

2. Cost reduction

Routine disputes may require fewer human resources.

3. Accessibility

Online dispute resolution can allow parties to participate without physically attending court.

4. Digital commerce

Digital assets, blockchain transactions and online contracts generate disputes that are naturally suited to digital platforms.

5. Cross-border transactions

Automated platforms can operate across geographical boundaries.

6. Large-volume disputes

High-volume claims involving standardised contracts may be suitable for automated processing.

5. From Human Adjudication to Algorithmic Decision-Making

Traditional adjudication follows approximately:

Claim → Evidence → Hearing → Legal analysis → Judgment → Enforcement

Automated resolution may become:

Digital transaction → Data capture → Algorithmic analysis → Automated determination → Digital execution

The major legal question is:

Can an algorithm legally perform a function that the law has assigned to a judge, arbitrator, or other authorised decision-maker?

The answer depends on the particular dispute-resolution regime and the authority granted to the system.

6. DIFC Digital Economy Court

The UAE's most concrete example of movement toward technologically enabled dispute resolution is the DIFC Digital Economy Court (DEC).

Part 58 of the DIFC Courts Rules specifically deals with Digital Economy Court claims.

It covers disputes involving areas such as:

fintech;

digital assets;

distributed ledger technology;

blockchain;

substantial databases;

artificial intelligence;

AI-dependent devices and systems.

The rules also permit digital conduct of proceedings and contemplate electronic dynamic systems using smart forms or AI-driven forms, including decision-tree software, to obtain information necessary for conducting and disposing of claims. (DIFC Courts)

This is an important development because it moves beyond merely putting paper processes online.

7. But Automated Forms Are Not the Same as Automated Judges

This distinction is crucial.

An AI-driven form can:

ask questions;

classify information;

identify missing information;

guide a claimant;

organise evidence;

route a dispute.

But this does not necessarily mean that AI has legal authority to issue a binding judicial judgment.

Therefore:

Automation of procedure ≠ automation of adjudication.

The present UAE trajectory is better described as:

human adjudication supported by increasingly sophisticated digital systems.

Current UAE dispute-resolution analysis likewise states that AI is presently used in an assistive rather than determinative capacity in the courts. (Global Practice Guides)

8. Automated Resolution and Civil-Law Principles

The shift must remain consistent with fundamental civil-law principles.

Important principles include:

A. Consent

Parties must know what dispute-resolution mechanism they have agreed to.

B. Legal capacity

The system cannot manufacture legal capacity where the underlying law does not recognise it.

C. Good faith

Automated contractual execution must not become a mechanism for abusive conduct.

D. Evidence

The underlying data must be authentic and reliable.

E. Due process

Parties must have a genuine opportunity to present their case where adjudication is involved.

F. Public policy

An automated result cannot be enforced merely because a computer produced it.

G. Judicial supervision

Where law requires judicial oversight, software cannot simply eliminate that requirement.

9. Automated Resolution and Electronic Evidence

Automation depends heavily upon data.

The UAE Evidence Law recognises electronic evidence and gives legal significance to electronically generated records where statutory conditions are satisfied. (UAE Legislation)

This creates a legal chain:

Data → Authentication → Admissibility → Weight → Decision

An algorithm can analyse data, but it cannot automatically make unreliable data legally reliable.

Example

Suppose an automated system determines:

"Buyer defaulted."

The legal question remains:

Was the transaction authentic?

Was the payment actually due?

Was payment received through another channel?

Was there a contractual grace period?

Was the system's timestamp accurate?

Was the data altered?

Was the algorithm correctly programmed?

Automation therefore does not eliminate evidentiary questions.

10. AI Errors and Human Responsibility

An important current UAE example is Stelian Gheorghe v BSA Ahmad Bin Hezeem & Associates LLP & Jimmy Haoula [2025] DIFC CFI 045.

The defendants argued that some of the claimant's material may have been generated using AI. The Court noted errors in the claim and evidence but did not treat AI as an autonomous legal decision-maker. The case illustrates that parties and their lawyers remain responsible for the accuracy and legal quality of material submitted to court. (DIFC Courts)

Principle

AI assistance does not transfer legal responsibility from the human litigant or lawyer to the machine.

11. Case Law 1 — Stelian Gheorghe v BSA Ahmad Bin Hezeem & Associates LLP [2025] DIFC CFI 045

Issue

Possible use of AI-generated material in court proceedings.

Principle

The court remained the human adjudicator. The existence of AI-generated material did not transform the legal process into automated adjudication.

Importance

The case demonstrates an important emerging rule:

Human parties remain responsible for what they place before a court, even when technology is used to prepare it.

This is highly relevant to future automated-resolution systems. (DIFC Courts)

12. Case Law 2 — Techteryx Ltd v Aria Commodities DMCC & Others [2025] DIFC DEC 001

This is a particularly significant UAE technology-law authority because it was heard in the DIFC Digital Economy Court.

The dispute involved digital-economy issues and major financial institutions. The case was transferred into the Digital Economy Court framework. (DIFC Courts)

Principle

The case demonstrates institutional recognition that technologically complex disputes can require a specialised judicial environment.

Importance for automated resolution

It illustrates a transitional model:

Traditional adjudication + specialised digital court + advanced technology

rather than:

AI replaces judge.

13. Case Law 3 — Naho v Neukirchi [2024] DIFC SCT 415

This case concerned electronic contracting and electronic signatures.

The DIFC Court considered the Electronic Transactions Law and explained the legal treatment of electronic signatures and electronic records. The court examined whether electronic communications could satisfy legal requirements relating to signing and contractual modification. (DIFC Courts)

Principle

Electronic processes can satisfy legal requirements where the applicable legislation recognises them.

Importance

Automated dispute resolution requires legally recognised digital transactions in the first place.

Thus:

Valid digital contract → valid digital record → potentially automatable contractual performance.

14. Case Law 4 — ICICI Bank Ltd v Bavaguthu Raghuram Shetty [2022] DIFC CFI 034

This case involved extensive examination of signatures, including electronically reproduced signatures.

The court considered how electronic or reproduced signatures should be evaluated and the evidentiary problems associated with transferring signatures between documents. (DIFC Courts)

Principle

The existence of a digital representation of a signature does not automatically resolve questions of authenticity.

Importance

This is crucial for automated resolution because algorithms often assume that the input data is authentic.

If:

Input = unreliable

then:

Automated conclusion = potentially unreliable.

15. Case Law 5 — Oheo Bank v Parker [2025] DIFC CA 006

This case concerned an application to set aside an arbitral award and the high threshold for court intervention.

The DIFC Court of Appeal examined the statutory grounds for setting aside and emphasised the limited nature of judicial intervention in arbitral awards. The proceedings themselves were conducted online, demonstrating how sophisticated technology can support adjudicative processes without eliminating the human tribunal. (DIFC Courts)

Principle

Technology can make arbitration highly digital while the legal decision remains an adjudicative function of the tribunal.

Relevance

This provides a useful distinction:

Digital arbitration is not necessarily automated arbitration.

16. Case Law 6 — Nihan v Nicholas & Niaz [2024] DIFC CA 012

This shareholder dispute involved arbitration and subsequent recognition and enforcement proceedings in the DIFC Courts.

The Court of Appeal dealt with the relationship between the arbitration process and court enforcement. (DIFC Courts)

Principle

Even where parties choose private arbitration, the legal system retains a role in recognition and enforcement.

Relevance to automation

An automated dispute mechanism cannot necessarily bypass the legal enforcement structure.

A smart contract may execute code, but legal enforcement can still require:

jurisdiction;

recognition;

enforcement;

public-policy review;

statutory supervision.

17. Case Law 7 — Najee v Nash [2023] DIFC SCT 496

The DIFC Small Claims Tribunal considered a dispute-resolution clause under which the parties had chosen arbitration.

The court treated arbitration as a distinct dispute-resolution mechanism and concluded that the existence of the arbitration agreement affected the court's jurisdiction. (DIFC Courts)

Principle

The agreed dispute-resolution mechanism matters.

Relevance

Automated resolution must therefore begin with a legally valid foundation.

A software platform cannot acquire jurisdiction merely because its programming says:

"This dispute will be resolved automatically."

There must be a valid legal basis for the mechanism.

18. Case Law 8 — Mirma v Mobal [2025] DIFC CA

The DIFC Court of Appeal dealt with a challenge to an arbitral award and the statutory framework governing judicial review of awards.

The case illustrates the distinction between:

an arbitral tribunal deciding the dispute;

a court supervising the legality of the arbitration.

(DIFC Courts)

Principle

Even a highly digital arbitration system remains embedded within a legal supervisory framework.

19. Blockchain and Automated Resolution

Blockchain can potentially change dispute resolution more radically than ordinary AI.

A blockchain-based system can record:

ownership;

transfers;

timestamps;

payment;

contractual performance;

digital-asset movements.

Smart contracts can then execute predetermined consequences.

Example

Contract:

Buyer pays AED 1 million by 1 October.

Smart contract:

If payment is received → release asset.

If payment is not received:

automatically impose contractual consequence.

This appears simple, but legal disputes arise where real-world facts differ from coded conditions.

20. The Oracle Problem

A smart contract cannot always observe the real world.

For example:

"Pay AED 500,000 if the construction project reaches 80% completion."

Who determines 80% completion?

A software system may require an oracle.

The oracle provides external information to the smart contract.

Therefore:

Real-world event → Oracle → Blockchain → Smart contract → Automated result

If the oracle is wrong, the automated system can produce the wrong outcome.

This creates a new legal question:

Who bears responsibility for an incorrect oracle?

21. Self-Executing Contracts and Civil Law

A smart contract may execute automatically, but legal validity remains a separate question.

There are two different concepts:

Technical execution

The computer code executed.

Legal performance

The parties legally performed their obligations.

These are not necessarily identical.

Formula

Code execution ≠ automatic legal validity

For example, a smart contract might automatically transfer a digital asset despite:

fraud;

mistake;

lack of authority;

incapacity;

breach of mandatory law;

invalid underlying contract.

A legal system must still determine what consequences follow.

22. The DIFC's Blockchain Development

The DIFC Courts have been exploring blockchain-based judicial infrastructure for years.

In 2018, the DIFC Courts and Smart Dubai announced work on a blockchain-based court initiative, including research into cross-border enforcement of judgments and future dispute-resolution mechanisms involving blockchain and smart contracts. (DIFC Courts)

More recently, the DIFC Courts announced specialised digital-custodian and blockchain-intelligence capabilities for appropriate complex cases. (DIFC Courts)

This demonstrates that the UAE's technological transformation is not limited to electronic filing.

23. Digital Economy Court Rules and AI

The DIFC Digital Economy Court rules are especially important.

They contemplate:

Digital proceedings

Proceedings can make extensive use of information technology.

Smart forms

Parties may provide information through dynamic electronic systems.

AI-driven forms

The rules expressly contemplate AI-driven forms and decision-tree software.

Digital assets

The Court can make orders concerning digital assets and digital access mechanisms.

Blockchain

Claims concerning blockchain and distributed-ledger technology fall within the Digital Economy Court's subject-matter framework.

(DIFC Courts)

This is probably the clearest current UAE example of a legal system being designed around digital-economy disputes.

24. Difference Between Automated Resolution and ODR

These concepts should not be confused.

Online Dispute Resolution (ODR)

ODR means the dispute is handled through an online platform.

It may still involve:

human mediator;

human arbitrator;

human judge.

Automated Dispute Resolution

The system itself performs substantial decision-making.

Therefore:

Every automated dispute-resolution system may be digital, but not every digital dispute-resolution system is automated.

25. Difference Between AI Assistance and AI Adjudication

26. Due Process Problem

Automated resolution creates a major procedural question:

Does a party have a right to know how the decision was reached?

Suppose an algorithm concludes:

"Claim rejected."

The claimant may ask:

Which evidence was rejected?

Which legal rule was applied?

Which facts were considered?

What weighting system was used?

Was the algorithm trained on biased data?

Can the decision be challenged?

Who is legally responsible?

These questions are much easier to answer when the decision is made by a human judge giving reasons.

27. Explainability

An automated legal decision should ideally be capable of explanation.

Example

Instead of:

"Risk score = 82; claim denied."

A legally meaningful system should potentially provide:

relevant facts;

applicable contractual provision;

relevant legal rule;

evidence relied upon;

reasoning;

conclusion;

appeal/review mechanism.

This produces:

Algorithmic result + legal reasoning + review mechanism

rather than merely:

Algorithmic result.

28. Bias and Equality

Automated systems can reproduce bias contained in:

training data;

historical judgments;

human classifications;

incomplete datasets;

poorly designed rules.

Therefore:

Automation does not automatically produce neutrality.

An algorithm trained on historically unequal outcomes may reproduce those outcomes.

The legal system therefore needs mechanisms for:

testing;

auditing;

validation;

monitoring;

correction;

human review.

29. Human Oversight

Human oversight is especially important where the dispute involves:

large financial claims;

family rights;

property;

employment;

personal data;

fraud;

public policy;

constitutional or statutory rights;

vulnerable parties.

A useful model is:

Human-in-the-loop

AI → Recommendation → Human review → Decision

rather than:

Human-out-of-the-loop

AI → Final legal result

The former is much closer to the present UAE trajectory.

30. Automated Enforcement

Automated enforcement creates another distinction.

Traditional model

Judgment → enforcement application → enforcement authority → execution

Automated model

Trigger → algorithm → automatic execution

The second model can be much faster.

However, it raises questions concerning:

wrongful execution;

mistake;

fraud;

reversal;

third-party rights;

insolvency;

public policy;

jurisdiction.

Therefore, automatic execution should not be assumed to eliminate legal remedies.

31. Can an AI Judgment Be Challenged?

If an AI system merely provides administrative assistance, ordinary judicial review remains available because the human judge makes the decision.

If an AI system were eventually authorised to make binding determinations, legislation would need to clarify:

who is the legal decision-maker;

whether the decision is a judgment or contractual determination;

how notice operates;

how evidence is considered;

how reasons are provided;

how errors are corrected;

who bears liability;

how appeals work;

how enforcement occurs.

Without these rules, automated decision-making could create significant uncertainty.

32. Arbitration and Automated Resolution

Arbitration provides a potentially easier route to experimentation because parties have greater procedural autonomy.

For example, parties could agree to:

online arbitration;

electronic evidence;

AI-assisted document review;

algorithmic damages calculations;

automated scheduling;

smart-contract evidence.

But the tribunal still has to comply with:

the applicable arbitration law;

the arbitration agreement;

due process;

procedural fairness;

mandatory law;

public policy.

Thus:

Arbitration can be technologically automated without becoming legally autonomous.

33. Public Policy Limitation

An automated outcome cannot automatically override public policy.

For example, imagine a smart contract that says:

"No court may interfere with the automated result."

Such a clause cannot necessarily exclude mandatory legal rules.

The legal system retains control over questions such as:

arbitrability;

illegality;

public policy;

mandatory statutory protections;

jurisdiction.

This is why the relationship between code and law is central to automated dispute resolution.

34. Data Protection

Automated resolution requires substantial personal and commercial data.

The system may process:

names;

identification information;

financial records;

contracts;

communications;

biometric information;

transaction histories;

behavioural information.

Therefore, automated resolution must be considered alongside UAE data-protection requirements.

The more data an algorithm processes, the more important become:

purpose limitation;

security;

access control;

confidentiality;

data minimisation;

lawful processing.

35. Cybersecurity Risk

Automated resolution creates a new category of legal risk:

The dispute-resolution infrastructure itself may be attacked.

Possible attacks include:

manipulation of evidence;

alteration of transaction data;

identity theft;

algorithmic manipulation;

smart-contract exploitation;

oracle manipulation;

denial-of-service attacks.

Therefore:

Digital justice requires digital security.

36. Liability for Automated Decisions

Suppose an automated system makes an incorrect decision.

Who is responsible?

Potential candidates include:

software developer;

system operator;

court/tribunal institution;

data provider;

AI vendor;

party that supplied incorrect data;

oracle provider.

This creates a major emerging civil-law issue.

Possible liability formula

Wrongful automated output + legally recognised duty + causation + damage = potential liability

The precise application depends on the relevant legal framework and contractual arrangements.

37. The Shift Is Better Described as "Augmentation"

The current UAE position is more accurately understood as:

Traditional model

Human judge + paper + physical hearing

↓

Digital model

Human judge + electronic evidence + virtual hearing

↓

AI-assisted model

Human judge + AI tools + automated analysis

↓

Advanced digital model

Human decision-maker + algorithmic recommendations + smart execution

↓

Possible future model

Legally authorised automated resolution

The UAE has clearly progressed through the earlier stages, while the final stage raises unresolved legal and institutional questions. Current UAE dispute-resolution analysis similarly characterises AI's present role as supportive rather than a replacement for judges. (Global Practice Guides)

38. Key Legal Risks

39. Six Major Principles from the Case Law

The cases discussed above collectively support these principles:

1. Technology does not automatically replace legal responsibility

Stelian Gheorghe demonstrates the continuing responsibility of parties and lawyers for material presented to court. (DIFC Courts)

2. Digital transactions can have legal validity

Naho v Neukirchi demonstrates statutory recognition of electronic signatures and records. (DIFC Courts)

3. Electronic evidence still requires legal evaluation

ICICI Bank v Shetty illustrates that electronically reproduced signatures can generate questions of authenticity and evidentiary reliability. (DIFC Courts)

4. Digital courts can handle sophisticated technological disputes

Techteryx illustrates the operation of the DIFC Digital Economy Court. (DIFC Courts)

5. Technology can support arbitration without replacing the tribunal

Oheo Bank v Parker demonstrates technologically conducted proceedings alongside continuing human arbitral and judicial functions. (DIFC Courts)

6. Automated/digital systems remain subject to legal jurisdiction and enforcement

Nihan v Nicholas & Niaz illustrates the continuing relationship between arbitration and judicial recognition/enforcement. (DIFC Courts)

40. Examination Formula

For an exam answer, remember:

A-D-A-R

A — Automation

What function is being automated?

D — Due Process

Can parties understand and challenge the result?

A — Accountability

Who is legally responsible for the decision?

R — Review

Can a human court or tribunal review the result?

Therefore:

Automated Resolution = Technology + Legal Authority + Due Process + Accountability + Review

Without these components, technical automation does not necessarily produce legally valid dispute resolution.

41. Conclusion

The UAE is moving from traditional adjudication toward technologically enhanced dispute resolution, but the present transformation should not be described as a complete replacement of judges by artificial intelligence.

The most significant developments are:

electronic evidence;

virtual proceedings;

digital case management;

AI-assisted legal processes;

Digital Economy Court procedures;

blockchain-related judicial infrastructure;

digital-asset dispute resolution;

AI-driven forms and decision-tree systems;

technology-assisted arbitration.

The DIFC Digital Economy Court is particularly significant because its rules expressly contemplate AI-driven forms and decision-tree software, while also permitting proceedings to be conducted using information technology. (DIFC Courts)

At the same time, cases such as Stelian Gheorghe, Naho, ICICI Bank v Shetty, Techteryx, Oheo Bank v Parker, and Nihan demonstrate that technology remains embedded within a framework of human responsibility, evidence rules, jurisdiction, arbitration law and judicial supervision. (DIFC Courts)

Core principle:

The UAE's emerging model is not simply "judges versus machines"; it is the integration of automation into a legally supervised system in which technology can increasingly perform procedural, analytical and execution functions while legal authority, accountability and review remain essential.