Civil Law And Distributed Artificial Intelligence Legal Collectives And Consensus Based Governance In Uae .
1. Introduction
Distributed Artificial Intelligence (DAI), legal collectives, and consensus-based governance represent an emerging area of civil law in which multiple AI systems, human participants, autonomous software agents, and distributed networks cooperate to make decisions, manage assets, perform contractual obligations, and regulate collective activities.
In the United Arab Emirates (UAE), this subject is particularly significant because of the country's developments in artificial intelligence, blockchain technology, digital assets, financial technology, and smart-government services.
Traditional civil law generally assumes that legal decisions are made by identifiable natural persons or legally recognised organisations. Distributed AI challenges this assumption because decisions may emerge from interactions between multiple autonomous systems without one central decision-maker.
Examples include:
Decentralised Autonomous Organisations (DAOs).
AI-managed investment collectives.
Blockchain-based commercial organisations.
Autonomous digital marketplaces.
Distributed contractual networks.
AI-powered voting and governance systems.
Token-based collective ownership arrangements.
Machine-to-machine commercial transactions.
Central legal question: When multiple AI agents and human participants collectively make decisions through a decentralised network, who possesses legal authority, who bears civil liability, and how can UAE law recognise and enforce the resulting decisions?
The UAE legal framework does not generally recognise an AI network or an informal digital collective as a separate legal person merely because it operates autonomously. Its legal status must be established through the applicable rules governing companies, foundations, contracts, digital assets, and other recognised legal structures.
2. Meaning of distributed artificial intelligence
Distributed Artificial Intelligence refers to systems in which several independent or semi-independent AI agents cooperate, exchange information, process data, and coordinate activities to achieve a common objective.
Unlike a conventional AI system controlled by one central operator, distributed AI may involve multiple agents operating across different computers, organisations, or geographical jurisdictions.
Main characteristics
Decentralisation: Decision-making is distributed among multiple participants or systems.
Autonomy: Individual AI agents can perform specified operations without continuous human intervention.
Collaboration: Different agents exchange information and coordinate tasks.
Consensus: Collective decisions may be reached through predefined voting or agreement mechanisms.
Transparency: Blockchain or distributed ledgers may record decisions and transactions.
Automation: Smart contracts may execute decisions automatically.
Collective governance: Participants may exercise voting or administrative powers through digital mechanisms.
Example
Imagine a UAE-based digital investment collective involving:
500 human members.
20 AI investment agents.
A blockchain-based voting system.
A smart contract controlling collective funds.
An automated dispute-resolution mechanism.
Members submit proposals, AI agents evaluate financial risks, participants vote, and smart contracts execute approved decisions.
This arrangement creates questions about contractual authority, ownership, responsibility, voting rights, and civil liability.
3. Meaning of legal collectives
A legal collective is an arrangement through which multiple individuals or entities coordinate their activities to pursue a common legal, economic, or organisational objective.
Traditional examples include:
Companies.
Partnerships.
Associations.
Foundations.
Cooperatives.
Joint ventures.
Investment funds.
Digital technology has created additional forms of collective organisation.
3.1 Decentralised Autonomous Organisations
Governance and DAOs. In the third lesson of the second… | by GREED Academy | Medium
What Is a DAO? Organisations for a Decentralised Age
12
A DAO is an organisation or governance arrangement in which blockchain-based rules and smart contracts coordinate activities, voting, and asset management.
Its principal characteristics include:
Digital membership.
Token-based voting.
Automated treasury management.
Collective proposals.
Distributed decision-making.
Smart-contract execution.
However, a DAO is a technological and organisational concept, not automatically a legally recognised form of company or legal person under UAE law.
3.2 AI-managed legal collectives
These are collective arrangements in which AI agents perform administrative or decision-support functions.
For example, an AI system might:
Evaluate membership applications.
Review financial proposals.
Recommend voting outcomes.
Monitor contractual compliance.
Identify conflicts of interest.
Allocate collective resources.
The legal question is whether these functions are merely administrative or whether they involve the exercise of legally significant authority.
3.3 Machine-based collective participation
A further development involves AI agents participating in collective processes as automated representatives of human members or organisations.
Such agents may submit proposals, negotiate commercial terms, or cast authorised votes.
Their authority must derive from a legally valid arrangement involving the persons or entities they represent.
4. UAE legal framework governing distributed AI and legal collectives
The UAE does not have one comprehensive federal statute governing every form of DAO or distributed AI collective. Instead, several legal regimes may apply.
4.1 Federal Decree-Law No. 25 of 2025 – Civil Transactions Law
The Civil Transactions Law, effective from 1 June 2026, provides the general private-law framework relevant to distributed AI arrangements.
Its relevance includes:
Legal personality.
Capacity.
Contract formation.
Obligations.
Ownership.
Civil responsibility.
Compensation.
Agency and representation.
General contractual relationships.
A distributed AI collective may create legally enforceable contractual relationships among its members even where the collective itself does not possess independent legal personality.
For example, members may agree to contribute capital, follow voting procedures, and distribute profits according to predefined rules.
The resulting contractual rights and obligations must be analysed under the applicable civil and commercial legislation.
4.2 Federal Decree-Law No. 32 of 2021 – Commercial Companies Law
The Commercial Companies Law is particularly important because it provides recognised legal structures for conducting collective commercial activities.
Depending on the arrangement, relevant structures may include:
Limited liability companies.
Private joint-stock companies.
Public joint-stock companies.
Partnerships.
Other recognised corporate structures.
A DAO conducting commercial activity in the UAE cannot assume that blockchain-based membership or token voting replaces the requirements for establishing and registering an appropriate legal entity.
The company may use AI and distributed governance technology while remaining subject to its statutory corporate structure.
4.3 DIFC Foundations Law and ADGM foundations framework
Foundations can be particularly relevant to digital collectives.
A foundation may provide a legal structure for holding assets, administering collective interests, and establishing governance arrangements.
The DIFC and ADGM have distinct legal frameworks for foundations.
These structures may be useful for certain blockchain projects, digital asset arrangements, and collective asset-management structures.
However, a foundation is not automatically equivalent to a DAO, and its governing documents must comply with the applicable law.
4.4 DIFC Law No. 2 of 2024 – Digital Assets Law
This legislation is important for understanding digital assets and property-related interests within the DIFC.
It provides a specialised framework for determining the legal treatment of digital assets.
Its relevance to distributed AI collectives includes:
Digital asset ownership.
Transfer of digital assets.
Custody and control.
Digital asset transactions.
Property-related disputes.
It is important to distinguish the DIFC's legal regime from the federal onshore UAE framework.
4.5 Federal Decree-Law No. 46 of 2021 – Electronic Transactions and Trust Services
This legislation supports electronic contracting and authentication.
It is relevant to:
Electronic membership agreements.
Digital voting records.
Electronic signatures.
Automated contractual communications.
Digital authorisation.
Electronic records of collective decisions.
A blockchain voting record may provide evidence of a decision, but its legal effect depends on the authority of the voters and the applicable contractual and statutory requirements.
4.6 Federal Decree-Law No. 45 of 2021 – Personal Data Protection
Distributed AI systems frequently process personal information belonging to members, customers, investors, and other participants.
The Personal Data Protection Law is relevant to:
Lawful processing.
Data minimisation.
Transparency.
Security.
Data-controller responsibilities.
Data-processor responsibilities.
Cross-border data transfers.
An AI collective cannot avoid applicable data protection obligations merely because information is distributed across several network participants.
4.7 Federal Decree-Law No. 34 of 2021 – Combating Rumours and Cybercrimes
This legislation is relevant where distributed systems are used for:
Unauthorised access.
Digital impersonation.
Fraudulent voting.
Manipulation of electronic records.
Unlawful data access.
Cyberattacks against governance systems.
The criminal and civil consequences must be assessed separately.
4.8 Virtual asset regulatory frameworks
The UAE has developed specialised regulatory frameworks for virtual assets.
Relevant authorities include:
Dubai Virtual Assets Regulatory Authority (VARA).
Securities and Commodities Authority.
Central Bank of the UAE.
Financial Services Regulatory Authority of ADGM.
Dubai Financial Services Authority of DIFC.
The relevant regulatory requirements depend on the activity, location, nature of the asset, and type of service being provided.
A token-based AI collective may therefore be subject to financial regulation in addition to ordinary civil and corporate law.
5. Legal personality of distributed AI collectives in the UAE
Legal personality is one of the most important issues in distributed AI governance.
A legal person can generally:
Own property.
Enter into contracts.
Sue and be sued.
Hold rights and obligations.
Appoint representatives.
Bear civil liability.
An AI collective may perform many of these functions technologically. However, technological functionality does not automatically create legal personality.
5.1 Can a DAO become a legal person in the UAE?
The answer is potentially yes, through an applicable recognised legal structure.
Two particularly important developments are the ADGM Distributed Ledger Technology Foundations Regulations 2023 and the Ras Al Khaimah DAO Associations Regime (DARe), unveiled in October 2024.
A. ADGM Distributed Ledger Technology Foundation
Separate legal personality
Under Section 3 of the ADGM DLT Foundations Regulations 2023, a registered DLT Foundation is a legal entity with separate legal personality.
It can be established to use, develop, facilitate, or support distributed ledger technology or issue tokens.
Its governance structure can include:
Founders.
Foundation Council.
Tokenholders.
Beneficiaries.
Guardian, where required.
Its Charter determines important governance arrangements, including delegated matters and tokenholder voting rights.
(1)_DLT Foundations Regulations 2023+1
B. RAK DAO Associations Regime
DAO-specific legal framework
The Ras Al Khaimah Digital Assets Oasis introduced its DAO Associations Regime in October 2024.
It provides a registration framework for qualifying decentralised autonomous organisations to operate through a recognised association structure.
This is important because it offers an institutional route for translating decentralised governance into a legally identifiable organisational arrangement.
Home Page+1
5.2 Legal personality of the AI itself
The position is different for the AI agents operating within these organisations.
An AI agent does not automatically become a legal person merely because it:
Votes on proposals.
Negotiates contracts.
Controls software.
Manages digital assets.
Executes transactions.
Participates in governance.
The legal entity or authorised persons operating the system remain central to determining rights and liabilities.
5.3 Legal personality versus technological autonomy
| Question | Legal position |
|---|---|
| Can an AI agent vote? | Yes, if authorised under the governance arrangements |
| Can an AI agent execute a smart contract? | Yes, technically and subject to legal validity |
| Can an AI agent own assets independently? | No automatic legal recognition |
| Can a registered DLT Foundation own assets? | Yes, subject to its applicable law |
| Can a DAO sue or be sued? | Depends on its recognised legal structure |
| Does autonomous operation eliminate human liability? | No |
6. Consensus-based governance and civil law
Consensus-based governance means that collective decisions are reached through a predefined agreement mechanism rather than exclusively through the authority of a central administrator.
In distributed systems, consensus may be achieved through:
Token-weighted voting.
Majority voting.
Supermajority voting.
Delegated voting.
Stake-based consensus.
AI-assisted voting.
Multi-agent agreement.
Human-machine consensus.
6.1 Token-based voting
Members receive governance tokens that provide voting rights.
For example, a UAE digital investment foundation may have 1,000 members.
Each member receives voting rights according to the governing Charter.
A proposal to invest AED 10 million in a digital infrastructure project is submitted.
The voting system determines whether the proposal satisfies the required threshold.
The legal question is whether the resulting decision was made by the correct persons, under the correct procedure, and within the authority delegated to them.
6.2 Majority consensus
A collective may require more than 50% of valid votes for a decision.
This can be used for:
Routine administrative decisions.
Budget approval.
Membership decisions.
Operational proposals.
However, a majority cannot necessarily override mandatory legislation or statutory protections.
6.3 Supermajority consensus
Certain decisions may require 66.67%, 75%, or another specified threshold.
Examples include:
Amending the governance Charter.
Dissolving the organisation.
Changing its principal objectives.
Transferring substantial assets.
Modifying tokenholder rights.
The applicable threshold should be established in the governing constitutional documents and remain consistent with mandatory legal requirements.
6.4 AI-assisted consensus
AI systems can evaluate proposals before human voting.
For example, five AI agents may independently assess:
Financial risk.
Legal compliance.
Cybersecurity.
Environmental impact.
Commercial feasibility.
Their recommendations may be combined into a single decision-support report.
However, the AI-generated recommendation should not automatically replace legally required approvals.
7. Legal validity of collective decisions
A central issue is whether a blockchain-based collective decision can create legally enforceable obligations.
The answer depends on several requirements.
7.1 Authority
The person or body approving the proposal must have the legal authority to do so.
7.2 Consent
The members must have agreed to the applicable governance arrangements.
7.3 Procedural compliance
Voting must comply with the required quorum, threshold, notice, and voting rules.
7.4 Legality
The decision must not violate mandatory law, public order, or applicable regulatory requirements.
7.5 Authenticity
The voting records and digital identities must be sufficiently reliable.
7.6 Enforceability
The decision must create a legal obligation recognised by the applicable legal framework.
Example: A DAO votes to distribute AED 2 million among its tokenholders. Although the blockchain records approval, distribution may still be challenged if the proposal violated the Charter, involved an unlawful transaction, or failed to obtain a mandatory regulatory approval.
8. Smart contracts as instruments of collective governance
DAO Crypto: How Decentralized Governance Works
Smart contracts are computer programs that automatically execute specified instructions when predetermined conditions are met.
In distributed AI collectives, they may govern:
Treasury management.
Voting procedures.
Membership rights.
Asset transfers.
Profit distribution.
Payment obligations.
Governance amendments.
8.1 Legal recognition
Federal Decree-Law No. 46 of 2021 supports electronic transactions and legally recognised electronic records and signatures.
However, a smart contract's technical execution must be distinguished from the legal validity of the underlying agreement.
8.2 Smart-contract mistakes
Suppose an AI agent submits an incorrect proposal and a smart contract transfers AED 500,000 to the wrong wallet.
Questions include:
Was the transaction authorised?
Was the software functioning as intended?
Was there an error in programming?
Who controlled the AI agent?
Was there negligence?
Can the transaction be reversed?
Who suffered the resulting loss?
The fact that blockchain transactions may be technically irreversible does not necessarily prevent a court from recognising a legal claim for restitution or compensation.
8.3 Conflict between code and law
A particularly difficult problem arises when a smart contract executes an action that conflicts with the organisation's legal obligations.
For example, a smart contract may distribute funds even though the transaction violates a regulatory restriction.
The legal obligation is not displaced merely because the computer code executed automatically.
9. Agency and representation in distributed AI governance
Agency is a fundamental civil-law concept.
In conventional transactions, a principal authorises an agent to act on their behalf.
Distributed AI creates a new form of agency-related problem.
Suppose a UAE company authorises an AI agent to negotiate supplier contracts.
The AI agent:
Identifies suppliers.
Negotiates prices.
Accepts quotations.
Generates contractual documentation.
Submits purchase orders.
The legal question is whether the AI agent acted within the authority granted by the company.
9.1 Actual authority
The principal expressly authorises specified actions.
9.2 Implied authority
Authority may arise from the nature and scope of the assigned responsibilities, subject to the applicable legal rules.
9.3 Excess of authority
If an AI agent exceeds its authorised limits, the principal's liability must be examined under the relevant agency, contract, and third-party reliance principles.
9.4 Apparent authority
A company that presents an AI system as authorised to transact may create questions concerning reasonable third-party reliance and responsibility for its representations.
The AI itself does not become an independent legal agent in the same way as a natural person. Its actions must be attributed through the relevant legal relationship.
10. Civil liability in distributed AI collectives
The allocation of liability is one of the most difficult aspects of distributed AI governance.
Federal Decree-Law No. 25 of 2025 is particularly important.
Article 245 provides that the general harmful-act liability framework applies to harmful acts committed by natural or legal persons, subject to special legislation.
Article 246 provides that an act causing harm to another creates an obligation to compensate.
Article 247 distinguishes direct harm from harm caused through causation.
These provisions are relevant when an AI collective causes financial, property, contractual, or other legally recognised damage.
UAE Legislation+1
10.1 Liability of developers
Developers may face claims where defective design, negligent programming, or other legally actionable conduct causes damage.
10.2 Liability of operators
Operators may be responsible where inadequate monitoring, deployment, security, or supervision causes harm.
10.3 Liability of governance members
Members may face exposure depending on their actual legal status, participation, authority, applicable entity structure, and the relevant liability rules.
Token ownership alone should not be treated as automatically establishing liability under UAE law.
10.4 Liability of the registered entity
Where the collective operates through a recognised legal person, the entity may bear contractual or civil liability for obligations attributable to it.
10.5 Liability of AI service providers
A provider may face contractual or other legal responsibility where its services fail to satisfy applicable obligations.
10.6 Liability of human decision-makers
Individuals who approve or implement harmful decisions may face responsibility where their own conduct satisfies the relevant legal requirements.
11. Fiduciary duties and collective governance
Collective governance must protect the interests of participants and the organisation.
This is particularly important when AI systems control collective funds or make decisions affecting tokenholders.
Relevant governance duties may include:
Good faith.
Loyalty.
Reasonable care.
Skill and diligence.
Avoidance of conflicts of interest.
Proper purpose.
Accountability.
Transparency.
In an ADGM DLT Foundation, the Regulations expressly establish duties for councillors and guardians.
These include acting in accordance with the Charter, exercising independent judgment, acting honestly and in good faith, and exercising reasonable care, skill, and diligence.
The framework also provides a guardian mechanism and gives the Foundation Council important supervisory powers.
DLT Foundations Regulations 2023 (March 31)+1
Example
An AI governance system recommends that the foundation invest its entire treasury in a project owned by one of its councillors.
The proposal receives majority tokenholder approval.
Even if the voting process is technically valid, the transaction may raise questions concerning conflicts of interest, proper purpose, fiduciary duties, and the council's statutory responsibilities.
Consensus does not automatically legitimise a conflicted or unlawful decision.
12. Data protection and privacy in distributed AI collectives
Distributed AI systems often require access to large volumes of personal information.
Examples include:
Member identities.
Voting histories.
Financial information.
Transaction records.
Behavioural profiles.
Biometric information.
Communications.
Investment preferences.
The UAE Personal Data Protection Law must be considered where the relevant processing falls within its scope.
Important principles include:
Lawful processing.
Purpose limitation.
Data minimisation.
Transparency.
Appropriate security.
Accountability.
Lawful cross-border transfers.
A decentralised network cannot avoid legal responsibility simply by distributing data among multiple nodes.
The organisation must determine who controls processing, who acts as a processor, and which participants have access to personal information.
13. Collective ownership and distribution of digital assets
Distributed AI collectives may own or administer assets collectively.
Examples include:
Cryptocurrency.
Tokenised investments.
Intellectual property.
Digital platforms.
Real estate interests.
Commercial receivables.
Investment portfolios.
The legal structure determines how those assets are owned.
Three possible arrangements are particularly important.
| Structure | Ownership arrangement |
|---|---|
| Registered company | Assets owned by the company |
| ADGM DLT Foundation | Assets held by the separate foundation |
| Contractual collective | Ownership and obligations determined by applicable agreements and property law |
A tokenholder does not automatically own a proportional share of every asset held by a DAO.
The token's legal rights must be determined by the relevant Charter, contractual documentation, and applicable legislation.
14. Important case laws: Distributed AI, DAOs and consensus-based governance
Jurisdictional clarification: There is no established UAE appellate case-law doctrine specifically recognising distributed AI agents as legal persons or deciding the civil liability of an AI-governed DAO. The following seven comparative judicial authorities are useful for understanding the legal questions that UAE courts may encounter. They are not binding UAE precedents.
Case 1. Commodity Futures Trading Commission v. Ooki DAO (2023)
Citation: No. 3:22-cv-05416, US District Court for the Northern District of California, 2023.
Facts:
Ooki DAO operated a decentralised financial platform through blockchain-based governance.
Its participants used governance tokens to participate in organisational decisions.
The Commodity Futures Trading Commission alleged that the organisation had operated unlawfully in connection with regulated commodity transactions.
A major issue was whether a decentralised organisation could be treated as an unincorporated association and held legally accountable.
Decision:
On 8 June 2023, the court entered default judgment against Ooki DAO.
The judgment included a monetary penalty of USD 643,542 and a permanent injunction.
The court's earlier order treated Ooki DAO as an unincorporated association capable of being sued under the applicable law.
daollc.org+1
Legal principles:
Decentralised governance does not automatically eliminate legal accountability.
An organisation operating through smart contracts may still be subject to legal proceedings.
Organisational conduct can create regulatory liability.
Token-based governance does not automatically provide immunity.
Legal classification depends on the applicable jurisdiction.
Relevance to UAE:
This case is particularly important for UAE-based DAOs.
A digital collective cannot assume that decentralisation removes its obligations under UAE law.
A properly registered legal structure can provide clarity regarding the entity responsible for contractual and regulatory obligations.
Case 2. Sarcuni v. bZx DAO (2023)
Citation: No. 3:22-cv-00618, US District Court for the Southern District of California, 27 March 2023.
Facts:
Users of the bZx decentralised finance protocol suffered losses following a cyberattack.
They brought proceedings against the DAO and associated participants.
The plaintiffs alleged negligence and argued that the governance-token holders constituted a general partnership.
The central issue was whether tokenholders exercising collective management powers could be treated as partners.
Decision:
At the motion-to-dismiss stage, the court held that the plaintiffs had pleaded sufficient facts to allege that a general partnership existed among the relevant governance-token holders.
The decision did not constitute a final finding that every tokenholder was liable.
The case was subsequently dismissed with prejudice.
daollc.org+1
Legal principles:
Collective control can have legal consequences.
Governance participation may be relevant to partnership classification.
Decentralised decision-making does not necessarily eliminate duties.
The economic and organisational substance of an arrangement matters.
Token-based voting may create liability questions beyond the software itself.
Relevance to UAE:
This case highlights the importance of identifying the actual legal relationship among participants.
A UAE collective should specify whether its members are investors, beneficiaries, shareholders, contractual participants, or authorised governance representatives.
Case 3. Samuels v. Lido DAO (2024)
Citation: No. 3:23-cv-06492, US District Court for the Northern District of California, 18 November 2024.
Facts:
An investor brought proceedings involving Lido DAO and several venture capital firms after suffering alleged losses associated with LDO tokens.
The plaintiff argued that Lido DAO constituted a general partnership.
The defendants challenged the legal characterisation of the organisation.
Decision:
The court concluded at the pleading stage that the plaintiff had alleged sufficient facts to support a reasonable inference that a general partnership had been formed.
The court considered allegations concerning founders, governance-token holders, and active participation in the organisation's management.
This was not a final determination of partnership liability.
The proceedings were subsequently stayed pending appeal, according to the reported procedural status in October 2026.
daollc.org+1
Legal principles:
The substance of collective management matters.
An organisation may be treated as a partnership despite lacking a conventional partnership agreement.
Governance-token participation can become relevant to legal classification.
Institutional investors may face questions concerning their actual participation in management.
Legal uncertainty should be addressed through appropriate organisational structuring.
Relevance to UAE:
The case illustrates why founders and investors should establish clear boundaries between passive investment and management authority.
It also demonstrates the importance of choosing a recognised legal structure rather than relying entirely on decentralised software.
Case 4. Quoine Pte Ltd v. B2C2 Ltd (2020)
Citation: [2020] SGCA(I) 2, Singapore Court of Appeal.
Facts:
Quoine operated a cryptocurrency exchange on which B2C2 conducted automated algorithmic trading.
Transactions were concluded through computer programs without direct human participation in each individual trade.
Following a technical malfunction, cryptocurrency transactions occurred at highly abnormal prices.
Quoine attempted to reverse the transactions, arguing that they were affected by mistake.
Decision:
The Singapore Court of Appeal rejected the principal argument that the trades were void because of unilateral mistake and upheld the contractual claim arising from their cancellation.
The court also examined the role of the programmers' knowledge when applying contractual mistake principles to algorithmic transactions.
elitigation.sg+1
Legal principles:
Algorithmically concluded transactions can create enforceable contracts.
Traditional contractual doctrines can apply to automated transactions.
Programmer knowledge may be relevant when assessing algorithmic conduct.
A malfunctioning platform can create contractual liability.
Automated systems do not remove the need for legal interpretation.
Relevance to UAE:
This case is important for distributed AI collectives because multiple autonomous agents may collectively negotiate or execute transactions.
The legal effect of those transactions must be evaluated through ordinary principles of consent, authority, contractual formation, and responsibility.
Case 5. Tulip Trading Ltd v. Bitcoin Association for BSV & Others (2023)
Citation: [2023] EWCA Civ 83, Court of Appeal of England and Wales.
Facts:
Tulip Trading claimed ownership of substantial Bitcoin holdings whose private keys had been lost or stolen.
It argued that certain Bitcoin developers exercised control over the network and should owe fiduciary duties to cryptocurrency owners.
The claimant sought assistance through changes to the software that would allow access to the assets.
The defendants disputed that they owed such duties.
Decision:
The Court of Appeal allowed the appeal to proceed, holding that it was arguable that developers could owe fiduciary duties to cryptocurrency owners in certain circumstances.
The court considered the developers' alleged control, discretionary decision-making powers, and relationship with users.
It did not finally determine that the developers owed all the duties alleged.
Bailii
Legal principles:
Decentralised systems can involve identifiable persons exercising significant control.
Software developers may have legally relevant powers.
Fiduciary duties may arise from the actual relationship and responsibilities undertaken.
Technical decentralisation does not necessarily eliminate human control.
The exercise of discretionary authority can create legal obligations.
Relevance to UAE:
This judgment is especially important where AI agents, protocol developers, and governance councils collectively manage digital property.
The legal system must identify who exercises real decision-making power rather than simply assuming that a decentralised network has no responsible actors.
Case 6. AA v. Persons Unknown (2019)
Citation: [2019] EWHC 3556 (Comm), High Court of England and Wales.
Facts:
A company suffered a ransomware attack and paid Bitcoin to unknown attackers.
The company subsequently sought a proprietary injunction to recover or protect the cryptocurrency.
The central question was whether Bitcoin could be legally treated as property.
Decision:
The court held that Bitcoin could constitute property and granted the relevant proprietary relief.
It adopted the reasoning that cryptocurrency could satisfy the established characteristics of property.
Legal principles:
Digital assets can qualify as property.
Intangible assets may attract proprietary protection.
Courts can grant remedies involving cryptocurrency.
Digital assets can be traced and recovered through appropriate legal procedures.
Relevance to UAE:
Distributed AI collectives frequently manage digital assets.
The case provides comparative support for treating digital assets as legally significant property rather than merely information stored on a computer network.
Bailii
Case 7. D'Aloia v. Persons Unknown and Others (2024)
Citation: [2024] EWHC 2342 (Ch), High Court of England and Wales.
Facts:
Fabrizio D'Aloia was deceived into transferring cryptocurrency to a fraudulent trading platform.
He attempted to trace the cryptocurrency through multiple blockchain transactions and sought recovery from a cryptocurrency exchange.
The proceedings involved difficult questions concerning cryptocurrency ownership, tracing, and constructive trusts.
Decision:
The court reaffirmed the property character of cryptocurrency but dismissed the claim against the exchange because the claimant failed to establish the necessary tracing connection to the assets received by that exchange.
The judgment demonstrated that recognising cryptocurrency as property does not automatically prove that a particular defendant received or holds the claimant's assets.
Bailii+1
Legal principles:
Digital assets can attract proprietary rights.
Blockchain transaction histories may assist asset tracing.
Ownership and control must be distinguished.
Constructive trust claims require appropriate legal and factual foundations.
Technical transaction records do not eliminate the burden of proof.
Relevance to UAE:
This judgment is relevant to disputes involving AI-managed collective treasuries, token transfers, digital asset custody, and unauthorised transactions.
Comparative analysis of the seven cases
| Case | Main legal issue | Importance |
|---|---|---|
| CFTC v. Ooki DAO | DAO accountability | Organisational liability |
| Sarcuni v. bZx DAO | Tokenholder relationships | Partnership exposure |
| Samuels v. Lido DAO | Collective management | Governance responsibility |
| Quoine v. B2C2 | Algorithmic contracting | Automated contractual liability |
| Tulip Trading | Developer authority | Fiduciary responsibility |
| AA v. Persons Unknown | Crypto property | Digital asset ownership |
| D'Aloia | Blockchain tracing | Recovery and proprietary remedies |
Overall judicial principle: The courts in these comparative jurisdictions have increasingly examined the actual legal and economic relationships behind decentralised technology. They have not accepted that smart contracts, autonomous algorithms, or token voting automatically remove traditional legal responsibilities.
15. Dispute resolution in distributed AI collectives
Dispute resolution is a particularly difficult issue because decentralised organisations may involve participants located in different countries.
Potential disputes include:
Disagreement over voting outcomes.
Unauthorised treasury transfers.
AI-generated financial losses.
Smart-contract malfunction.
Membership disputes.
Misuse of governance tokens.
Conflicts of interest.
Breach of collective agreements.
15.1 Internal dispute resolution
A collective may establish internal procedures for handling disagreements.
These can include:
Governance review committees.
Independent auditors.
Human review of AI decisions.
Tokenholder appeals.
Independent technical investigations.
15.2 Arbitration
An appropriately drafted arbitration agreement may provide for the resolution of disputes involving members, service providers, or the legal entity.
The UAE's arbitration framework, including Federal Law No. 6 of 2018 on Arbitration, as amended, is relevant where its jurisdictional and contractual requirements are satisfied.
15.3 Court proceedings
Where a registered legal person exists, it can participate in court proceedings according to the applicable legal framework.
15.4 On-chain dispute resolution
Some DAOs use automated voting or specialised decentralised arbitration protocols.
However, an on-chain decision does not necessarily replace a legally binding judgment or arbitral award.
The legal enforceability of an internal resolution depends on the governing agreement, jurisdiction, and applicable law.
16. Practical hypothetical cases under UAE civil law
The following examples demonstrate how distributed AI governance could generate civil disputes in the UAE.
Hypothetical 1: AI-governed investment collective
A UAE-based DAO has 2,000 tokenholders and collectively manages AED 50 million.
Its AI system evaluates investment proposals and recommends transactions.
A majority vote approves an investment that subsequently causes substantial financial losses.
Legal analysis:
The court may need to examine:
Whether the DAO has separate legal personality.
Whether the investment was properly authorised.
Whether the AI system operated within its mandate.
Whether the governance council fulfilled its duties.
Whether the members were passive investors or active managers.
Whether negligence or contractual breach occurred.
The AI system itself does not automatically become the liable legal person.
Hypothetical 2: AI agent executes an unauthorised contract
A UAE company authorises an AI agent to negotiate contracts up to AED 100,000.
Due to a software error, the AI agent concludes a contract worth AED 2 million.
Legal analysis:
The dispute may involve:
Limits of actual authority.
Apparent authority.
Electronic contractual formation.
Third-party reliance.
Negligence.
Contractual interpretation.
Responsibility for system deployment.
The company's liability will depend on the authority communicated to the third party and the applicable legal rules.
Hypothetical 3: Manipulation of DAO voting
A participant controls multiple digital wallets and uses them to manipulate a collective governance vote.
The proposal results in the transfer of AED 5 million from the collective treasury.
Legal analysis:
Potential issues include:
Fraudulent voting.
Unauthorised transactions.
Breach of governance rules.
Cybercrime.
Restitution.
Compensation.
Liability of persons controlling the relevant wallets.
A blockchain record showing a successful vote does not necessarily establish that the vote was legally valid.
Hypothetical 4: AI agents disagree over collective decisions
A UAE-based legal foundation operates ten AI agents.
Five agents recommend approving a commercial contract, while the other five reject it.
The system automatically resolves the disagreement through a majority algorithm, but the contract creates a substantial financial loss.
Legal analysis:
The dispute may require examination of:
The foundation's Charter.
The authority delegated to the AI agents.
The voting algorithm.
The supervisory role of the council.
The existence of human oversight.
The contractual consequences of the decision.
The fact that an automated consensus mechanism selected the outcome does not necessarily establish that the decision was properly authorised.
Hypothetical 5: Cyberattack on an AI collective treasury
Hackers compromise the governance system of a registered digital foundation and transfer cryptocurrency to unauthorised wallets.
Legal analysis:
The foundation may seek:
Injunctive relief.
Asset preservation.
Digital asset tracing.
Restitution.
Compensation.
Disclosure and forensic orders where legally available.
The responsibility of custodians, cybersecurity providers, developers, and administrators must be assessed individually.
Hypothetical 6: Conflict between AI governance and the Foundation Council
An ADGM DLT Foundation allows tokenholders to approve proposals through blockchain voting.
A proposal is approved that conflicts with the foundation's Charter.
The Foundation Council refuses to implement the decision.
Legal analysis:
The relevant questions include:
Whether the proposal fell within delegated matters.
Whether the voting procedure complied with the Regulations.
Whether the decision was lawful.
Whether the council exercised its powers properly.
Whether tokenholders have an enforceable right to implementation.
This illustrates that decentralised voting must operate within the recognised legal governance structure.
17. A proposed model for AI collective governance in the UAE
A balanced framework should combine technological decentralisation with identifiable legal responsibility.
Proposed UAE distributed AI governance architecture
1. Human members and founders
Establish objectives, rights and governance rules
2. Recognised legal entity
Company, foundation or qualifying DAO association
3. Distributed AI agents
Analyse proposals, risks and contractual requirements
4. Consensus and human approval
Voting, verification and authorised decision-making
5. Smart-contract execution
Automated performance of authorised transactions
6. Audit and legal accountability
Records, liability, remedies and dispute resolution
Essential safeguards
The proposed system should include:
Legal registration and identification of the responsible entity.
Clearly defined AI authority.
Human oversight for high-impact decisions.
Independent cybersecurity audits.
Transparent voting procedures.
Conflict-of-interest controls.
Traceable digital records.
Financial risk management.
Appropriate insurance.
Accessible dispute-resolution mechanisms.
18. Proposed legal reforms for the UAE
Although the UAE has developed important legal structures for blockchain-based foundations and DAOs, additional clarification would help address the intersection of AI and decentralised collective governance.
18.1 Legal recognition of AI governance systems
Legislation could clarify the circumstances in which AI systems may act as authorised decision-support or execution mechanisms for recognised legal entities.
This would not require granting AI independent legal personality.
18.2 Distributed collective liability framework
A clearer framework could distinguish among:
Founders.
Governance councils.
Active tokenholders.
Passive investors.
AI developers.
AI operators.
Custodians.
Technology service providers.
This would improve the allocation of responsibility.
18.3 Mandatory AI governance documentation
High-impact AI collectives could be required to maintain:
AI system descriptions.
Decision-making records.
Authority limitations.
Risk assessments.
Human oversight procedures.
Incident reports.
18.4 Smart-contract legal standards
Regulatory guidance could address:
Code verification.
Contractual interpretation.
Execution errors.
Reversal mechanisms.
Emergency suspension.
Unauthorised transfers.
Dispute procedures.
18.5 Tokenholder protection
A clearer framework could establish minimum standards for:
Voting rights.
Disclosure.
Treasury transparency.
Conflicts of interest.
Minority protection.
Governance amendments.
Exit rights.
18.6 AI accountability insurance
Insurance mechanisms could cover certain losses arising from:
AI system errors.
Cybersecurity failures.
Unauthorised automated transactions.
Negligent system deployment.
Technology-provider failures.
18.7 Cross-border governance standards
Because distributed AI systems operate internationally, the UAE could further develop cooperation concerning:
Digital asset recovery.
Cross-border evidence.
Recognition of foreign judgments.
International arbitration.
Regulatory information exchange.
19. Critical evaluation of the UAE legal position
The UAE framework presents three important developments.
First: Institutional recognition of digital collectives.
The ADGM DLT Foundations Regulations and RAK DAO Associations Regime provide routes for qualifying decentralised organisations to obtain recognised legal structures.
Second: Continued application of traditional civil-law principles.
The Civil Transactions Law provides a framework for contractual obligations, legal responsibility, and compensation even when the underlying transaction involves AI or blockchain technology.
Third: No automatic legal personality for AI agents.
The legal status of the organisation must be distinguished from the technological autonomy of its software.
There are also three continuing challenges:
| Legal challenge | Underlying problem |
|---|---|
| Attribution | Identifying who legally made or authorised an AI decision |
| Accountability | Determining who bears responsibility for collective harm |
| Enforceability | Reconciling blockchain consensus with mandatory legal rules |
A particularly important point is that decentralisation and legal decentralisation are not necessarily the same thing. A network may distribute its technical operations while a relatively small group of founders, developers, operators, or council members retains significant practical control.
Legal analysis should therefore examine actual power, authority, and responsibility rather than relying exclusively on the system's technological description.
20. Important legislative authorities for academic research
| Legislation | Legal significance |
|---|---|
| Federal Decree-Law No. 25 of 2025 | General civil obligations and liability |
| Federal Decree-Law No. 32 of 2021 | Commercial company structures |
| Federal Decree-Law No. 46 of 2021 | Electronic transactions and authentication |
| Federal Decree-Law No. 45 of 2021 | Personal data protection |
| Federal Decree-Law No. 34 of 2021 | Cybercrime regulation |
| Federal Law No. 6 of 2018 | Arbitration |
| ADGM DLT Foundations Regulations 2023 | Legal personality and DLT governance |
| RAK DAO Associations Regime 2024 | DAO association registration |
| DIFC Law No. 2 of 2024 | Digital asset legal framework |
An important regulatory qualification is that UAE AI governance is still developing. There is no generally applicable standalone federal AI statute comprehensively regulating all distributed AI collectives. The existing framework combines general legislation, sector-specific rules, free-zone regimes, and non-binding AI governance guidance.
CMS Law+2
21. Conclusion
Distributed Artificial Intelligence, legal collectives, and consensus-based governance are transforming the traditional understanding of civil-law relationships.
In conventional civil law, contractual decisions, organisational authority, and property administration are generally associated with identifiable individuals and legally recognised organisations.
Distributed AI introduces a different model in which multiple autonomous agents, human participants, blockchain protocols, and automated governance mechanisms cooperate to produce legally significant decisions.
The UAE has taken important steps towards accommodating this development through its Civil Transactions Law, Commercial Companies Law, electronic transactions legislation, ADGM DLT Foundations Regulations, and RAK DAO Associations Regime.
The comparative cases of Ooki DAO, Sarcuni, Samuels v. Lido DAO, Quoine v. B2C2, Tulip Trading, AA v. Persons Unknown, and D'Aloia demonstrate that decentralised technology does not automatically eliminate contractual, proprietary, fiduciary, or regulatory responsibilities.
The most important future legal development will be the creation of a balanced framework that recognises technological autonomy while preserving legal accountability.
Ultimately, the central principle should be that decision-making may be distributed, but legal responsibility must remain identifiable and enforceable.
For the UAE, the objective should not necessarily be to grant independent legal personality to AI agents. It should be to develop legally recognised collective structures capable of using distributed AI while maintaining transparency, accountability, property protection, contractual certainty, and effective civil remedies.

comments