1. Introduction

Artificial Intelligence (AI) Contract Governance in Thailand concerns the legal rules, contractual obligations, rights, liabilities and enforcement mechanisms governing agreements that are created, negotiated, reviewed, performed or terminated through artificial intelligence technologies.

AI is increasingly used in commercial transactions, banking, insurance, procurement, employment, logistics, digital marketplaces and international trade. Businesses use AI-powered systems to negotiate prices, generate contractual clauses, evaluate contractual risks, approve transactions and automatically execute contractual obligations.

These developments raise important civil-law questions:

Can an AI system legally conclude a contract?

Who is responsible when AI generates an incorrect contractual term?

Can an AI-generated electronic agreement be enforced?

Who bears liability when an autonomous system accepts an unauthorised offer?

Can a party cancel a contract because an AI system made an operational mistake?

How should contractual disputes involving AI-generated evidence be resolved?

Thailand primarily addresses these issues through its existing Civil and Commercial Code, Electronic Transactions Act B.E. 2544 (2001), Personal Data Protection Act B.E. 2562 (2019), Consumer Protection Act B.E. 2522 (1979) and other applicable legislation.

Important legal position: Thailand does not yet have a comprehensive AI-specific statute in force. Consequently, AI contractual relationships are principally governed by existing civil and electronic transaction laws.

Dej-Udom & Associates+2

2. Legal Framework Governing AI Contracts in Thailand

A. Civil and Commercial Code

The Thai Civil and Commercial Code (CCC) provides the foundation for contractual relationships involving AI.

Its general principles apply whether an agreement is negotiated through traditional communication or an AI-powered platform.

Important provisions include:

Application to AI contracts

Suppose a Thai company deploys an AI chatbot to negotiate supply agreements.

The chatbot offers 10,000 products at an incorrectly calculated price because of a programming defect.

The legal questions include:

Whether the chatbot's communication constituted an offer.

Whether acceptance created a binding agreement.

Whether the company authorised the system to communicate contractual terms.

Whether the company can rely on a mistake or another defence.

Whether damages are recoverable.

The legal consequences depend on the actual transaction, the authority given to the system and the applicable contract principles.

B. Electronic Transactions Act B.E. 2544 (2001)

This is one of the most important statutes governing AI-enabled contracts in Thailand.

The legislation recognises electronic communications and electronic contractual transactions.

Important provisions include:

Section 7 – Recognition of electronic data

Legal validity cannot be denied merely because information exists in electronic form.

AI-generated contractual communications may therefore be legally relevant electronic records.

Section 8 – Written evidence

Electronic information can satisfy legal requirements for written documents if it remains accessible and can be reproduced without alteration of its meaning.

Section 9 – Electronic signatures

Electronic methods can satisfy signature requirements when they identify the relevant person and indicate that person's approval, subject to the statutory reliability requirements.

Section 13 – Electronic contract formation

Offers and acceptances may be communicated electronically. A contract cannot be denied legal effect merely because the offer or acceptance was made electronically.

Section 13/1 – Automated electronic communications

The provision addresses offers made through electronic communications and automated-response systems, including circumstances in which an electronic offer constitutes an invitation to make an offer.

Section 13/2 – Automated contract formation

A contract concluded between an automated message system and a natural person, or through interactions between automated message systems, cannot be denied validity or enforceability merely because no natural person participated in each individual automated act.

Section 17/1 – Input errors

The provision provides a mechanism for withdrawing an erroneous portion of an electronic communication in specified circumstances where an individual makes an input error when interacting with another party's automated system.

etda.or.th+1

Legal significance: Section 13/2 is particularly important because it expressly recognises automated contract formation without requiring human participation in every individual transaction.

However, it does not make AI a separate legal person. The relevant contractual rights and obligations ordinarily remain attributable to the persons or businesses operating the systems.

C. Personal Data Protection Act B.E. 2562 (2019)

AI contract systems frequently process:

Customer identities.

Financial information.

Employment information.

Consumer preferences.

Biometric information.

Contractual transaction histories.

The PDPA establishes obligations concerning lawful processing, transparency, security and data-subject rights.

Sections 19, 23, 24, 26 and 37 are particularly relevant.

For example, a company using AI to assess customer creditworthiness must examine whether the personal data is lawfully collected and processed, whether an appropriate legal basis exists and whether the required safeguards are maintained.

D. Consumer Protection Act B.E. 2522 (1979)

AI-powered platforms may generate standard contractual terms for consumers.

Examples include:

Automated subscription agreements.

AI-generated insurance terms.

Online purchase agreements.

Automated cancellation provisions.

Algorithmically determined prices.

Consumer protection requirements remain applicable.

Businesses cannot assume that AI-generated contractual clauses are automatically enforceable simply because customers accepted them electronically.

E. Draft Artificial Intelligence Act

Thailand's ETDA released a revised Draft Artificial Intelligence Act for public consultation in July 2026.

The proposal addresses AI risk classification, regulatory oversight, transparency and liability.

However, it remains a draft rather than enacted legislation. Its proposed obligations must not be treated as existing statutory requirements.

Baker McKenzie+2

3. Meaning and Scope of Artificial Intelligence Contract Governance

AI contract governance involves controlling the complete contractual lifecycle.

AI contractual governance lifecycle

Contract creation

AI prepares contractual documents, commercial terms and draft clauses.

Contract negotiation

AI communicates proposals, evaluates counteroffers and negotiates prices.

Contract acceptance

Automated systems communicate acceptance and establish contractual relationships.

Contract execution

AI systems initiate payments, deliver services or trigger contractual performance.

Contract monitoring

AI identifies deadlines, compliance failures and contractual risks.

Dispute resolution

Parties examine electronic records, system logs, contractual authority and liability.

4. Major Legal Issues in AI Contract Governance

A. AI-Generated Contract Formation

AI can independently communicate contractual proposals through automated platforms.

For example, a Thai logistics company may use an AI procurement system to negotiate delivery charges with suppliers.

The system might:

Identify available suppliers.

Propose delivery prices.

Negotiate payment terms.

Accept qualifying proposals.

Generate purchase orders.

Under Section 13/2 of the Electronic Transactions Act, the resulting contract cannot be denied validity merely because automated systems participated without human intervention in each individual act.

Nevertheless, the parties must establish the identity of the contracting entities, the system's authority and the relevant contractual terms.

B. AI Contractual Authority and Agency

A significant question concerns whether an AI system can act as an authorised representative of a company.

For example, a company authorises an AI platform to negotiate purchases up to THB 500,000.

The AI system mistakenly accepts an agreement worth THB 800,000.

The dispute may involve:

The scope of authority.

The company's representations to the supplier.

The supplier's knowledge of any transaction limits.

The contractual conditions governing acceptance.

The consequences of an unauthorised transaction.

Traditional agency principles under the Civil and Commercial Code may become relevant.

AI itself does not ordinarily become an agent with independent legal personality. Rather, the authority and conduct of the persons operating or deploying it must be examined.

C. AI Contractual Mistakes

AI systems sometimes produce inaccurate contractual information because of:

Incorrect training data.

Faulty programming.

Model hallucinations.

Inaccurate pricing calculations.

Misinterpretation of contractual instructions.

Suppose an AI system mistakenly inserts a payment obligation of THB 5 million instead of THB 500,000.

The parties may dispute whether the agreement reflects their genuine intention, whether a legally relevant mistake occurred and whether the contract can be avoided.

The relevant provisions include Sections 154 and 171 of the CCC, together with the rules governing contractual interpretation and formation.

D. AI Contractual Breach

AI systems may fail to perform contractual functions.

Examples include:

An AI procurement system failing to issue purchase orders.

Automated payment systems failing to transfer funds.

AI scheduling software missing delivery deadlines.

Contract-management software incorrectly terminating agreements.

Liability depends on the contractual promises, the responsible party's obligations, causation and any applicable contractual defences.

Section 420 of the CCC may also become relevant where the conduct independently constitutes a wrongful act.

E. AI Contractual Liability Allocation

A well-drafted AI contract should identify responsibility among:

These are potential contractual responsibilities, not automatic legal liabilities. Their application depends on the agreement and applicable legislation.

F. AI Contractual Evidence

Electronic evidence is particularly important in AI contractual disputes.

Relevant records include:

AI-generated communications.

Electronic signatures.

Chat histories.

API transaction records.

System logs.

Automated acceptance notifications.

Version histories.

Payment records.

The Electronic Transactions Act and applicable evidence rules govern the legal significance and admissibility of these materials.

5. Important Thai Case Laws

7 relevant Supreme Court decisions

Important qualification: Thailand has not yet developed an established body of Supreme Court decisions specifically deciding disputes over contracts concluded by generative AI or autonomous AI agents. The following decisions concern electronic communications, electronic signatures, contractual evidence and related principles. They provide relevant legal foundations for future AI contract disputes, but should not be misrepresented as AI-specific judgments.

Case 1: Supreme Court Decision No. 8089/2556

Subject: Electronic signatures and electronic loan transactions

The dispute concerned a cash-credit card transaction.

The borrower used a cash card and personal identification number to withdraw money and confirmed the transaction.

The Supreme Court recognised that:

Entering a personal identification number could be comparable to signing electronically.

The electronic transaction could establish evidence of borrowing.

Electronic records could satisfy relevant written-evidence requirements.

Legal principle

Electronic authentication can establish legally relevant evidence of a contractual transaction.

Relevance to AI contract governance

An AI platform that authenticates users and records their contractual approvals may produce legally significant evidence, provided the statutory requirements are satisfied.

Case 2: Supreme Court Decision No. 6757/2560

Subject: Electronic communications and discharge of debt

The dispute concerned a creditor who communicated through Facebook that the debtor did not need to repay the outstanding amount.

The Supreme Court recognised the electronic communication as written evidence of the creditor's intention to discharge the debt.

The decision demonstrates that:

Electronic messages can constitute legally recognised communications.

Digital messages may satisfy written-evidence requirements.

Electronic expressions of intention can have substantive civil-law consequences.

Relevance to AI contract governance

AI-generated messages may become relevant to contractual amendments, waivers and debt settlements.

However, the party's intention and authority must still be established.

Case 3: Supreme Court Decision No. 123/2565

Subject: Electronic communication and contractual acceptance

This decision concerned an extension of the redemption period relating to a sale with right of redemption.

The legal issue involved whether electronic communications established the necessary written evidence and acceptance.

The absence of an appropriate response meant that the communication could not be treated as an established electronic agreement satisfying the relevant requirements.

Legal principle

A communication alone does not necessarily establish contractual acceptance.

Relevance to AI contract governance

An AI system cannot automatically treat silence or an unanswered proposal as acceptance unless the applicable legal rules and contractual circumstances support that conclusion.

Case 4: Supreme Court Decision No. 1112/2566

Subject: LINE communications and loan documentation

The dispute involved electronic communications exchanged through LINE concerning financial transactions.

The Court considered whether the messages constituted sufficient written evidence of a loan under Section 653 of the CCC and the Electronic Transactions Act.

The electronic communications were not sufficient to establish the necessary signed written loan evidence in the circumstances of the case.

Legal principle

The existence of an electronic conversation does not automatically prove the existence of an enforceable loan agreement.

The content, identity of the sender and legal significance of the communication matter.

Relevance to AI contract governance

AI-generated contractual conversations must contain sufficiently clear terms and reliable attribution. Merely producing a chat transcript does not establish every contractual element.

Case 5: Supreme Court Decision No. 2085/2566

Subject: Electronic communications and discharge of debt

This dispute concerned a creditor's WeChat messages indicating that the debtor no longer needed to repay the debt.

The Supreme Court considered the electronic communications under Section 340 of the CCC and Sections 7 and 8 of the Electronic Transactions Act.

The messages were accepted as important evidence of the creditor's intention to discharge the debt.

Legal principle

Electronic communications may establish legally effective expressions of contractual intention.

Relevance to AI contract governance

An AI-powered debt-management system might communicate settlement offers or debt waivers.

Businesses must ensure that such systems have appropriate authority and that their communications accurately reflect the intended legal position.

Case 6: Supreme Court Decision No. 2162/2567

Subject: LINE messages as written evidence of a loan

Particularly important

This decision is especially relevant to digital contract formation.

The dispute concerned multiple transfers of money and LINE conversations in which the lender described the proposed loans and requested the borrower to type "ตกลง" (meaning "agreed").

The Supreme Court distinguished between transactions.

The first transfer lacked sufficient written loan evidence.

Transfers 2–18 were supported by messages identifying the loan arrangements and the borrower's electronic acceptance.

Transfers 19–20 lacked the necessary acceptance messages.

The Court recognised the relevant LINE communications as written evidence bearing the borrower's electronic signature under Sections 7, 8 and 9 of the Electronic Transactions Act.

Legal principle

An electronic communication can satisfy contractual written-evidence and signature requirements when its contents, attribution and context sufficiently establish the relevant agreement.

Relevance to AI contract governance

This decision provides an important foundation for AI-negotiated transactions.

For example, an AI procurement system could communicate:

The contractual price.

The quantity of goods.

The delivery obligations.

The payment conditions.

An authenticated electronic acceptance could then provide evidence of contractual agreement.

LawNode+2

Case 7: Supreme Court Decision No. 3622/2568

Subject: LINE messages and insufficient evidence of a loan

This decision provides an important counterbalance to Decision No. 2162/2567.

The dispute concerned a transfer of THB 100,000 and electronic conversations between the parties.

The Court found that the LINE conversation did not contain sufficient information establishing that the transferred money constituted a loan.

The electronic communication's existence was not enough to establish the necessary contractual evidence under Section 653 of the CCC.

Legal principle

Electronic evidence must establish the substance of the underlying contractual transaction. A bank transfer and a digital conversation do not automatically prove a loan.

Relevance to AI contract governance

An AI platform must preserve sufficiently clear contractual terms.

A transaction record showing that money was transferred cannot, by itself, establish every contractual obligation.

LawNode+1

6. Comparative Analysis of the Seven Cases

Overall judicial direction: Thai Supreme Court decisions recognise the legal significance of electronic contractual communications, but require adequate evidence of intention, identity, agreement and the transaction's actual substance.

7. AI Contract Governance in Different Commercial Sectors

A. Banking and Financial Services

AI systems are used for:

Digital lending agreements.

Automated loan approvals.

Credit-risk assessments.

Payment agreements.

Debt restructuring.

The main legal concerns include incorrect financial calculations, unauthorised transactions, data protection and electronic evidence.

B. E-Commerce and Digital Marketplaces

AI platforms can negotiate product prices and process customer orders.

Contractual disputes may concern:

Incorrect product descriptions.

Automated pricing errors.

Unauthorised discounts.

Subscription renewals.

Automated cancellation.

The CCC, Electronic Transactions Act and consumer protection legislation may apply.

C. Employment Contracts

Employers may use AI to prepare employment agreements, evaluate applicants and administer contractual obligations.

Potential disputes include:

Incorrect salary provisions.

Automated termination decisions.

Discriminatory contractual conditions.

Improper employee monitoring.

Confidentiality breaches.

The Labour Protection Act B.E. 2541 (1998) and applicable employment laws remain relevant.

D. Insurance Contracts

AI can assess applications, generate insurance quotations and assist with claims processing.

Legal issues include inaccurate risk assessments, misleading policy terms and automated claim rejection.

The applicable insurance legislation and regulatory requirements must be considered alongside general contract law.

E. International Commercial Contracts

Thai businesses may enter into international agreements through AI-powered negotiation systems.

Important issues include:

Applicable law.

Jurisdiction.

Arbitration.

Electronic signatures.

Cross-border data processing.

Recognition of electronic records.

Contractual authority.

Parties should expressly establish governing law and dispute-resolution mechanisms.

8. Essential Clauses for AI-Enabled Contracts in Thailand

Businesses should consider incorporating the following contractual provisions.

Example of an AI Authority Clause

A contractual provision could state:

The AI-enabled contractual management system shall operate solely within the authority expressly granted by the contracting organisation. No transaction exceeding the prescribed financial or operational limits shall become authorised through the system without the required approval. The organisation shall maintain appropriate authentication, monitoring and electronic transaction records.

This is a suggested contractual drafting approach, not a statutory clause.

9. Major Challenges in Thailand's AI Contract Governance

1. Absence of AI-specific contractual legislation

Existing contract law was not originally designed specifically for autonomous AI systems.

2. Unclear responsibility

Multiple parties may participate in developing, deploying and operating the same AI platform.

3. Algorithmic opacity

It may be difficult to determine why an AI system generated a particular contractual term.

4. Electronic identity verification

Businesses must establish who authorised and accepted contractual communications.

5. Contractual mistakes

AI-generated mistakes can create disputes about intention, acceptance and contractual enforceability.

6. Personal data protection

AI platforms may process large quantities of confidential and personal information.

7. Cross-border contractual disputes

AI services may involve developers, cloud providers and customers located in different countries.

8. Uncertainty concerning future AI regulation

Thailand's proposed AI legislation may introduce additional regulatory requirements if enacted.

10. Recommendations for Strengthening Thailand's AI Contract Governance

Thailand should consider the following reforms.

A. Clear statutory recognition

Develop clearer guidance concerning contractual authority and automated decision-making.

B. AI contractual accountability

Establish transparent methods for allocating responsibility between developers, providers and deploying organisations.

C. Electronic evidence standards

Encourage reliable preservation of AI communications, authentication records and transaction histories.

D. Human supervision

Require appropriate human oversight for particularly significant financial and commercial transactions.

E. Consumer safeguards

Strengthen protection against misleading AI-generated contractual terms and unfair automated practices.

F. Data governance

Promote compliance with the PDPA and appropriate security standards.

G. Dispute-resolution mechanisms

Develop specialised procedures for disputes involving autonomous transactions, algorithmic decisions and AI-generated evidence.

H. Regulatory coordination

Encourage cooperation between ETDA, the Ministry of Digital Economy and Society, the Office of the Personal Data Protection Committee and relevant sectoral regulators.

11. Hypothetical Judicial Case Study

Case: AI-Generated Commercial Supply Agreement

Suppose a Thai manufacturing company operates an AI procurement platform.

The platform negotiates with a supplier and automatically accepts an offer for 20,000 industrial components.

However, the AI system mistakenly records the agreed price as THB 100 per unit instead of THB 1,000.

The supplier subsequently demands payment at the higher price.

The manufacturing company refuses, arguing that the AI made a mistake.

Questions before the court:

Did the AI system have authority to accept the supplier's offer?

Did the electronic acceptance create a binding agreement under Section 13/2?

Was the price error legally relevant under Sections 154 and 171 of the CCC?

Did the supplier know or have reason to know about the error?

Who bears responsibility for the system's defective operation?

What do the electronic records establish about the parties' intentions?

Possible legal outcome

The court would need to examine the complete contractual circumstances, including the parties' communications, authority arrangements, electronic records and applicable mistake provisions.

The fact that AI generated the acceptance would not, by itself, invalidate the agreement.

Equally, the fact that an electronic acceptance occurred would not necessarily prevent a party from relying on an applicable legal defence.

12. Conclusion

Artificial Intelligence Contract Governance in Thailand represents an important development in modern civil and commercial law.

Thailand already has a legal foundation for recognising electronic agreements and automated contractual transactions through the Civil and Commercial Code and the Electronic Transactions Act.

In particular, Section 13/2 of the Electronic Transactions Act provides express recognition of contracts concluded through automated messaging systems without requiring human participation in every individual act.

The Supreme Court decisions discussed above, particularly Decisions Nos. 2162/2567 and 3622/2568, demonstrate that electronic contractual evidence must establish the actual substance of the agreement and cannot be assessed merely by looking at the existence of a digital communication.

The future development of Thai AI contract governance will therefore depend on balancing technological innovation with contractual certainty, transparency, accountability, consumer protection and effective civil remedies.

The central legal principle is that automation does not eliminate contractual responsibility. Businesses and individuals remain subject to applicable contractual and statutory obligations even when AI systems perform important parts of the contracting process.

13. Important Legislation and Case Law for Academic Reference

Principal legislation

Thai Civil and Commercial Code.

Electronic Transactions Act B.E. 2544 (2001), as amended.

Personal Data Protection Act B.E. 2562 (2019).

Consumer Protection Act B.E. 2522 (1979).

Product Liability Act B.E. 2551 (2008).

Labour Protection Act B.E. 2541 (1998).

Computer Crime Act B.E. 2550 (2007).

Principal judicial decisions

Supreme Court Decision No. 8089/2556.

Supreme Court Decision No. 6757/2560.

Supreme Court Decision No. 123/2565.

Supreme Court Decision No. 1112/2566.

Supreme Court Decision No. 2085/2566.

Supreme Court Decision No. 2162/2567.

Supreme Court Decision No. 3622/2568.

Research note: These decisions provide relevant electronic-contract and civil-law precedents, not seven direct AI-contract judgments. Their precise application to AI disputes remains a matter for future judicial interpretation. No external website links have been included.