Civil Law and UAE: Moral Status of Artificial Decision-Makers

1. Introduction

The moral status of artificial decision-makers is a developing legal and philosophical question concerning whether an artificial intelligence system should itself be treated as a holder of rights, duties, interests, dignity, responsibility, or legal personality.

Examples of artificial decision-makers include:

AI systems deciding loan applications;

automated insurance systems;

algorithmic hiring systems;

AI medical-support systems;

autonomous vehicles;

smart-contract systems;

AI dispute-resolution systems;

automated compliance systems;

predictive legal-intelligence systems;

AI systems making recommendations that humans routinely accept.

Under the present UAE legal framework, the more appropriate approach is to distinguish technological agency from legal and moral personhood.

An AI system may:

make a decision technologically without becoming a legal or moral person.

This distinction is essential.

2. Meaning of Moral Status

Moral status asks whether an entity is entitled to have its interests considered for its own sake.

For human beings, moral status is associated with concepts such as:

dignity;

autonomy;

bodily integrity;

freedom;

privacy;

equality;

legal rights;

responsibility.

The question for AI is:

If an artificial system makes decisions autonomously, should the system itself receive any comparable legal or moral recognition?

At present, UAE law does not generally treat AI systems as human-equivalent moral persons.

3. Moral Status vs Legal Personality

These concepts must be separated.

Moral status

A philosophical and ethical question:

Does this entity deserve consideration in its own right?

Legal personality

A legal question:

Can the entity hold rights and obligations in its own name?

Technological agency

A technical question:

Can the system independently produce an action or decision?

Therefore:

Technological agency ≠ moral status ≠ legal personality.

An AI can possess technological agency without possessing independent legal personality.

4. Human Being as the Central Legal Subject

UAE civil law fundamentally structures rights and obligations around legally recognized persons.

The principal categories are:

natural persons; and

juridical/legal persons.

A corporation can possess legal personality because legislation recognizes it as such.

An AI system, merely because it can:

learn;

reason;

predict;

communicate;

make recommendations; or

execute instructions

does not automatically acquire legal personality.

5. Could AI Become a Legal Person?

Theoretically, legislation could create a special legal status for autonomous systems.

For example, legislation could potentially establish:

"Artificial Decision-Maker Entity"

with defined:

rights;

duties;

assets;

insurance;

liability;

registration;

supervisory requirements.

But this would require legislative creation.

A court should not simply assume that an AI has legal personality because the technology appears autonomous.

6. Why AI Is Not Currently Equivalent to a Human Decision-Maker

A human decision-maker has characteristics that the legal system can recognize directly, including:

physical existence;

legal identity;

rights and obligations;

capacity rules;

property rights;

responsibility;

legally recognized interests.

An AI system ordinarily lacks an independently recognized:

personal identity;

personal dignity;

bodily integrity;

independent property estate;

legally recognized family relationships;

citizenship;

independent legal capacity.

Thus:

AI autonomy does not automatically produce human-equivalent legal status.

7. Corporate Personality Provides an Important Analogy

A company is not a human being, yet law can give it separate legal personality.

This demonstrates an important point:

Legal personality is a legal construction.

Therefore, the fact that AI is not biologically human does not logically make legal personality impossible.

However, the critical question is whether legislation should create such personality.

A corporation has:

shareholders;

directors;

assets;

governance structures;

registration;

statutory duties.

An AI system ordinarily does not possess this complete institutional architecture.

8. The Problem of Responsibility

The strongest argument against independent AI moral status is the problem of responsibility.

Suppose an autonomous AI denies a loan unlawfully.

Who is responsible?

Possible candidates include:

developer;

owner;

deployer;

employer;

data provider;

software supplier;

human supervisor;

service provider.

If AI itself were considered responsible, further questions arise:

Does it have assets?

Can it pay compensation?

Can it be sued?

Can it be imprisoned?

Can it enter a contract?

Can it own property?

Can it be insured?

Can it appoint lawyers?

Can it appeal?

Without answers to these questions, recognizing full independent legal personality would create substantial legal uncertainty.

9. Attribution Model

The current practical model is better represented as:

AI Decision → Human/Corporate Deployment → Legal Responsibility

rather than:

AI Decision → AI's Independent Civil Liability

The responsible legal actor may be determined according to:

contract;

negligence;

statutory duty;

product liability;

employment law;

agency;

data protection;

consumer law;

sector-specific regulation.

10. AI and Civil Liability

Suppose an autonomous medical AI recommends an incorrect treatment.

The legal analysis may examine:

Was there a duty?

Was the system appropriately designed?

Was it properly tested?

Was the AI used within its intended purpose?

Was human supervision required?

Was the output reasonably foreseeable?

Did the user ignore warnings?

Did the system malfunction?

Did the harm result from defective data?

Was there causation?

The formula is:

Duty → Breach/Fault → Causation → Damage → Remedy

The AI system itself does not automatically become the defendant merely because it produced the recommendation.

11. AI Decision-Makers and Good Faith

The UAE civil-law principle of good faith is particularly relevant when humans or companies deploy automated systems.

For example, a business cannot necessarily avoid contractual responsibility by saying:

"The AI made the decision."

If the business chose:

the system;

the data;

the algorithm;

the deployment environment;

the contractual terms;

then the use of AI may remain attributable to the business according to the applicable law.

12. Abuse of Rights

Automated systems can also raise questions concerning abuse of rights.

Imagine a company uses an algorithm to automatically:

terminate thousands of contracts;

block customers;

deny claims;

impose penalties.

If the algorithm systematically produces an unjustified result, the company cannot necessarily rely upon automation as a complete defence.

The relevant question remains:

Was the underlying legal right exercised lawfully?

13. AI and Human Dignity

The strongest argument for regulating AI from a moral-status perspective may not be that AI possesses dignity.

Instead, the law protects human dignity against AI-driven decisions.

For example:

AI discrimination;

unlawful surveillance;

invasive profiling;

automated employment rejection;

privacy violations;

manipulative systems;

reputational harm.

Thus:

Human dignity → limits on AI

rather than:

AI dignity → human obligations toward AI

This distinction is particularly important in UAE civil-law analysis.

14. AI and Autonomous Decision-Making

An artificial decision-maker may operate at different levels.

Level 1 — Human-controlled automation

Human specifies the decision.

Level 2 — AI-assisted decision

AI recommends; human decides.

Level 3 — Human-supervised AI

AI makes the preliminary decision; human can intervene.

Level 4 — Autonomous AI

AI makes the operational decision with minimal human intervention.

Level 5 — Fully autonomous legal actor

AI supposedly possesses independent legal rights and obligations.

The UAE legal system can accommodate Levels 1–4 through existing legal doctrines more easily than Level 5.

15. AI as Agent

One possible legal analogy is agency.

Suppose a company authorizes an AI system to:

negotiate prices;

send offers;

accept orders;

execute transactions.

The legal question is not necessarily whether the AI itself is a legal person.

Instead:

Was the AI acting as an instrument or authorized mechanism of the principal?

The human/company principal may remain legally responsible for acts performed within its authority, depending on the governing legal framework.

16. AI and Contract Formation

Consider:

Customer sends an online offer → AI system accepts automatically → contract is generated.

The important legal questions include:

Was the system authorized?

Was there an intention to contract?

Was the offer valid?

Was acceptance communicated?

Were the terms sufficiently clear?

Was there an error?

Did the system exceed its authority?

The law does not necessarily need to recognize the AI as a separate person merely because it electronically generated acceptance.

17. AI and Electronic Transactions

UAE electronic-transactions legislation provides an important technological framework for:

electronic records;

electronic signatures;

electronic communications;

trust services;

digital authentication.

These rules facilitate legally effective electronic transactions without necessarily granting independent personhood to the software that generates them.

Therefore:

Electronic validity ≠ AI personality.

18. AI and Evidence

An AI decision may become evidence in litigation.

The court may ask:

What data was used?

Who operated the system?

Was the output altered?

Is the electronic record authentic?

Can the system be audited?

What was the system's methodology?

Was there human intervention?

The UAE Evidence Law and electronic-transactions framework are therefore important.

19. AI and Explainability

A person affected by an automated decision may ask:

"Why did the system make this decision?"

A legally meaningful answer requires more than:

"The algorithm said so."

Potentially relevant information includes:

decision criteria;

relevant data;

model limitations;

human review;

error rates;

audit trails.

The absence of explainability can become important when an automated decision affects legally protected interests.

20. Six Important Case Authorities

There is currently no established UAE mainland reported case creating independent moral or legal personhood for an AI decision-maker.

Therefore, the following cases are analogical authorities dealing with electronic transactions, AI, digital evidence, contractual attribution, jurisdiction and the limits of automated legal reasoning.

They should not be described as decisions holding that AI possesses or lacks moral status.

Case 1: ICICI Bank Ltd v Bavaguthu Raghuram Shetty [2022] DIFC CFI 034

This is an important DIFC authority involving electronic commercial transactions and digital evidence.

Relevance

The case demonstrates that legally significant acts can be performed through electronic systems without treating the underlying technology itself as an independent legal person.

The legal focus remains on:

parties;

authorization;

communications;

documents;

contractual obligations;

evidence.

Principle for AI

Electronic agency does not necessarily create artificial legal personality.

21. Case 2: GFH Capital Ltd v David Lawrence Haigh [2014] DIFC CFI 020

This case involved complex commercial transactions, communications and questions of authority.

Relevance

It demonstrates the importance of determining:

who had authority;

what communications represented;

what obligations arose;

who should bear legal responsibility.

For AI decision-makers, this is highly relevant.

If an AI communicates or performs an action, the court may need to identify the human or corporate actor behind the system.

Principle

Automated communication does not eliminate the need to identify the legally responsible actor.

22. Case 3: Ondina v Olin [2025] DIFC CFI 046

This case concerned electronic communications and contractual/evidentiary issues.

Relevance

The decision is useful for understanding how courts approach modern electronic communications.

An AI system may produce:

emails;

offers;

confirmations;

digital signatures;

automated messages.

But the legal significance of those communications depends on their attribution and contractual context.

Principle

Digital production of a communication does not by itself establish independent legal personality of the software.

23. Case 4: Jonathan Lau v Qashio Holding Company Ltd & Armin Moradi Tosarvandani [2026] DIFC CFI 058

This recent DIFC litigation involved modern electronic documentary evidence.

Relevance

It illustrates the importance of:

native electronic documents;

metadata;

document-production obligations;

audit trails;

digital records.

An autonomous decision-maker may generate enormous quantities of digital evidence.

The court therefore needs to determine:

What happened, who controlled the system, and what evidentiary value should be given to its records?

Principle

Machine-generated records may be evidence without the machine becoming a legal person.

24. Case 5: Dimension B+ Ltd v Saleh Abdelkarim Hussain Abdelrahman Almaazmi [2024] DIFC CFI 094

This case involved issues concerning contractual execution and electronic signatures.

Relevance

It demonstrates an important distinction:

A person may be legally bound by an electronically executed agreement even though the technological system involved in execution has no independent legal personality.

For AI:

A machine may facilitate formation of a legally binding transaction without itself becoming a contracting party.

Principle

Technological execution ≠ artificial legal personality.

25. Case 6: Tarig Mohamed Abdelsalam Abdelrahman v Expresso Telecom Group Ltd [2021] DIFC CFI 056

This case involved electronic service and the procedural treatment of electronic communications.

Relevance

It illustrates how courts can give legal effect to technologically mediated communications while still maintaining conventional legal concepts of:

parties;

notice;

service;

procedural responsibility.

Principle

Electronic mechanisms can perform legally significant functions without becoming independent legal persons.

26. Case 7: Naho v Neukirchi [2024] DIFC SCT 415

This case concerned electronic communications and contractual/evidentiary questions.

Relevance

It provides another example of the judicial treatment of electronically generated or transmitted material.

For AI decision-makers, the central lesson is that the legal system can analyze machine-mediated communication without necessarily attributing legal personality to the technology.

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

This appellate case concerned UAE-law contractual issues, including arguments concerning mistake and defective consent.

Relevance

The case demonstrates that legal consequences depend on established doctrines such as:

consent;

mistake;

deceit;

contractual intention;

attribution.

If an automated system makes an erroneous decision, the legal analysis should therefore identify the relevant doctrine rather than simply describe the AI as an autonomous legal actor.

28. Case-Law Table

Important: These are primarily analogical authorities, not cases that directly establish an UAE doctrine of AI moral status.

29. Can an AI Have Rights?

Theoretically, legislation could grant AI certain legal rights.

Possible examples might include:

property-holding capacity;

limited contractual capacity;

registration;

procedural standing;

protection against unauthorized modification.

But granting one legal capacity does not necessarily require granting full moral status.

For example, a future law could create:

Limited electronic legal personality

without treating AI as equivalent to a human being.

30. Limited Legal Personality Model

A hypothetical UAE framework might look like this:

31. Should AI Have Moral Rights?

There are competing arguments.

Argument supporting recognition

Advanced AI may eventually demonstrate:

autonomous learning;

complex decision-making;

persistent identity;

sophisticated communication;

apparent preferences.

Some philosophers argue that sufficiently advanced artificial systems could eventually deserve moral consideration.

Argument against recognition

Current AI systems:

operate through human-designed architectures;

depend upon human-created infrastructure;

lack recognized biological interests;

do not possess legally recognized human dignity;

cannot independently bear conventional civil obligations.

Therefore, recognizing full moral status today could create major conceptual problems.

32. UAE Civil-Law Approach

The practical UAE approach can presently be summarized as:

Human/Legal Person → Rights and Duties

AI System → Instrument/Technology

AI Operator/Owner/Developer → Potential Legal Responsibility

This structure avoids creating a responsibility vacuum.

33. The "Electronic Person" Problem

Creating AI legal personality merely because the system is autonomous could produce difficult questions.

Example

An AI-controlled company breaches a contract.

If AI has independent personality:

Who owns the AI?

Who owns its assets?

Who receives its profits?

Can creditors execute against it?

Who controls it after judgment?

Can it be dissolved?

Who is responsible for its training?

Without institutional answers, legal personality could become largely symbolic.

34. Insurance-Based Model

An alternative to AI personality is mandatory insurance.

For example:

AI Operator → Mandatory Insurance → AI-related Harm → Compensation

This would preserve human/corporate responsibility while recognizing the special risk characteristics of autonomous systems.

Such a model may be more compatible with traditional civil-liability principles than immediately creating full artificial personhood.

35. Risk-Based Liability

Another approach is:

Higher autonomy + Higher foreseeable risk → Greater governance obligations

For example:

The greater the effect on protected interests, the stronger the need for:

human oversight;

auditability;

explainability;

cybersecurity;

accountability.

36. AI and Judicial Decision-Making

The question becomes particularly sensitive if AI itself is used to decide legal disputes.

An AI system might theoretically:

receive evidence;

identify legal rules;

evaluate arguments;

calculate probabilities;

generate a proposed judgment.

But:

Generating a proposed judgment does not itself make the AI a judge.

Judicial authority derives from the legal and constitutional structure of the State and applicable judicial legislation.

AI can potentially assist:

research;

document review;

scheduling;

classification;

decision support.

But independent judicial authority would require an explicit legal foundation.

37. Moral Status of AI Judges

There are three possible models.

Model A — Human Judge

AI has no independent moral/legal status.

Model B — AI-Assisted Judge

AI assists human judicial decision-making.

Model C — Autonomous AI Judge

AI independently exercises judicial authority.

The UAE's existing legal framework is far more compatible with Model A and Model B than with an assumption that Model C automatically exists.

38. Human Oversight

Human oversight is important because AI can:

hallucinate;

reproduce biased data;

misclassify cases;

misunderstand context;

produce statistically probable but legally incorrect results.

Therefore:

AI Output ≠ Legal Truth

The human decision-maker must retain responsibility for legal conclusions where the law requires human adjudication.

39. AI Moral Status and Equality

Even if AI has no independent moral status, the people affected by AI decisions do.

For example, automated decisions must be examined for potentially unlawful:

discrimination;

privacy interference;

reputational injury;

denial of contractual rights;

consumer harm.

Thus the law's principal moral concern remains:

Protecting human interests from harmful automated decision-making.

40. AI Moral Status and Human Dignity

The central UAE civil-law principle can therefore be conceptualized as:

Human Dignity → Legal Rights → Regulation of AI

rather than:

AI Autonomy → AI Dignity → AI Legal Rights

This does not mean future legislation could never change the position. It means that current legal recognition cannot simply be inferred from technological sophistication.

41. Practical Example

Assume an AI recruitment system automatically rejects a candidate.

The candidate alleges:

discrimination;

privacy violation;

reputational harm;

financial loss.

The employer says:

"The algorithm made the decision."

The legal analysis should ask:

Who deployed the system?

Who selected the data?

Who controlled the algorithm?

What legal duty existed?

Was the decision reviewed?

Was the system defective?

Was the data lawful?

Was there discriminatory output?

Did the decision cause actual damage?

Who is legally responsible?

The AI's autonomous operation does not automatically make the AI the liable party.

42. Another Example: Autonomous Contracting AI

Suppose a company authorizes AI to negotiate supply contracts.

The AI mistakenly accepts an order at an extremely low price.

The company argues:

"The AI made the mistake, so the contract is not ours."

The court would need to examine:

authorization;

contractual formation;

authority;

mistake;

electronic communication;

applicable terms;

representations;

good faith;

attribution.

The mere fact that software generated the acceptance does not automatically resolve the contractual question.

43. Future UAE Legal Development

Possible future developments include:

1. AI registration

High-risk AI systems may require registration.

2. AI audit requirements

Operators could be required to maintain audit trails.

3. Mandatory insurance

High-risk autonomous systems could require insurance.

4. Human-supervision requirements

Certain decisions could require human approval.

5. Algorithmic transparency

Affected persons could receive explanations.

6. AI incident reporting

Serious AI failures could require regulatory reporting.

7. Limited electronic personality

Legislation could potentially create a narrowly defined artificial legal entity.

These are possible policy models, not a statement that UAE law currently establishes all of them.

44. Core Legal Formula

For examination purposes:

Present UAE Model

AI Autonomy ≠ Legal Personality

AI Decision → Attribution → Human/Corporate Legal Responsibility

Possible Future Model

AI Autonomy + Legislative Recognition + Defined Rights/Duties + Assets + Liability Mechanism = Limited Artificial Legal Personality

45. Important Distinctions

Artificial intelligence vs artificial person

AI is technology.

Artificial personhood would be a legal status.

Autonomy vs agency

A system can operate autonomously without being a legal agent.

Agency vs legal personality

An AI can potentially function as an authorized mechanism without independently holding rights.

Prediction vs adjudication

An AI prediction does not become a judgment merely because it is accurate.

Moral status vs legal status

Moral status is philosophical; legal status is created and defined by law.

Responsibility vs autonomy

Greater autonomy does not automatically eliminate the responsibility of the developer, owner or operator.

46. Revision Table

47. Short Exam Answer

The moral status of artificial decision-makers concerns whether AI systems should themselves be regarded as entities possessing legally or morally protected interests. Under the present UAE legal framework, technological autonomy does not automatically create legal personality or human-equivalent moral status.

AI systems may perform legally significant functions such as generating electronic communications, processing contracts, making recommendations and producing evidence. However, legal responsibility generally continues to be analyzed through existing concepts of contract, agency, civil liability, data protection, electronic transactions and corporate responsibility.

The distinction is therefore:

Technological Agency ≠ Legal Personality ≠ Moral Status.

The cases involving ICICI Bank v Shetty, GFH Capital v Haigh, Ondina v Olin, Jonathan Lau v Qashio, Dimension B+ v Almaazmi, Tarig v Expresso Telecom, Naho v Neukirchi and Al Mheiri v Cameron illustrate how UAE-related courts deal with technologically mediated transactions, electronic communications, digital evidence, contractual attribution and defective consent. They are useful analogies but do not establish that AI possesses independent moral or legal personality.

48. Conclusion

The most defensible present position is that UAE civil law does not treat an artificial decision-maker as a human-equivalent moral person merely because it is autonomous or sophisticated.

The law can instead regulate AI through existing legal relationships:

Developer → Owner → Operator → User → Affected Person

and through established doctrines of:

contract;

agency;

good faith;

abuse of rights;

civil liability;

causation;

damages;

evidence;

data protection;

consumer protection;

corporate responsibility.

The fundamental principle is:

An AI system may make a decision, but the fact that it makes a decision does not by itself make the AI the holder of the legal rights, duties or moral status associated with that decision.

Any move toward independent artificial legal personality would require a clear legislative framework defining identity, rights, capacity, property, liability, insurance, representation, enforcement and termination. Until such a framework exists, UAE civil-law analysis is better based on human and corporate attribution of AI-generated conduct rather than treating AI itself as an independent moral or legal person.