1. Introduction

Moral AI liability litigation in Thailand concerns civil disputes arising when artificial intelligence systems make, recommend, or facilitate decisions that create moral, ethical, social, financial, or personal harm.

The subject examines whether developers, AI operators, businesses, institutions, data controllers, and other responsible parties should be legally accountable when AI systems:

Discriminate against individuals or vulnerable groups.

Make unfair employment or recruitment decisions.

Produce harmful medical recommendations.

Manipulate consumers through behavioural profiling.

Violate personal dignity, privacy, or autonomy.

Generate defamatory or misleading information.

Make decisions that cause financial or psychological harm.

Use personal data without an appropriate legal basis.

Make decisions without adequate human supervision.

Cause harm through biased or unethical automated decision-making.

Thailand does not currently have a separate comprehensive civil-liability statute specifically establishing a legal personality or independent moral responsibility for artificial intelligence.

Instead, moral AI liability is primarily examined through existing civil-law principles, consumer protection, data protection, electronic transactions, constitutional values, and applicable sector-specific regulations.

Central legal principle: AI systems are not automatically independent legal persons responsible for their decisions. Civil responsibility generally needs to be attributed to legally recognised persons or organisations according to their conduct, duties, and involvement.

2. Legal framework governing moral AI liability in Thailand

A. Civil and Commercial Code

The Thai Civil and Commercial Code (CCC) provides the principal foundation for moral AI liability litigation.

Section 420 – General tort liability

Section 420 establishes liability where a person wilfully or negligently unlawfully injures another person's life, body, health, liberty, property, or rights.

It may apply to AI-related misconduct involving:

Discriminatory AI recruitment systems.

Negligently designed automated decision systems.

AI-generated defamatory statements.

Unlawful automated financial decisions.

AI systems causing foreseeable personal injury.

Negligent implementation of AI in professional services.

For example, if a company deploys a recruitment algorithm that negligently discriminates against qualified applicants and unlawfully damages their rights, Section 420 may provide a basis for a civil claim, depending on the facts and applicable legal protections.

Section 421 – Abuse of rights

This provision addresses the exercise of rights in a manner intended solely to cause injury to another person.

It may become relevant where an organisation uses its contractual or technological powers to deliberately disadvantage individuals through automated systems.

Section 425 – Employer liability

An employer may be jointly liable for wrongful acts committed by employees in the course of employment.

This may apply when employees negligently operate, supervise, or deploy AI systems in their professional activities.

Section 437 – Liability concerning dangerous things

This provision may become relevant to certain AI-enabled physical systems, depending on whether the relevant system falls within its statutory scope and the circumstances of the injury.

Examples may include autonomous machinery, AI-controlled vehicles, and automated industrial equipment.

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

The PDPA is one of the most important legal instruments for moral AI liability.

AI systems frequently rely on personal data for training, profiling, predictions, and automated recommendations.

Relevant provisions include:

These provisions may become relevant when AI systems process personal information unlawfully, use sensitive information improperly, or expose individuals to harm through inadequate data protection.

Importantly, the PDPA should not be interpreted as creating a universal right to human review of every AI-generated decision. The particular statutory rights and circumstances must be examined.

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

AI-powered commercial systems may create consumer disputes through:

Manipulative recommendation algorithms.

Misleading AI-generated advertisements.

Unfair automated pricing.

Deceptive chatbot representations.

Unclear automated subscription arrangements.

False claims about AI capabilities.

Businesses may face liability under consumer-protection legislation where their commercial conduct violates applicable legal requirements.

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

The Liability for Damages Arising from Unsafe Products Act B.E. 2551 may apply where an unsafe product causes legally recognised damage.

AI-enabled devices and software-integrated products require examination of the statutory definition of a product and the circumstances in which the injury occurred.

A developer's involvement in software design does not automatically establish liability under this Act.

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

This Act recognises electronic transactions and electronic records.

It is relevant to AI systems that:

Automatically accept contractual offers.

Generate electronic agreements.

Execute business instructions.

Communicate contractual terms.

Make automated transactions.

Its provisions help determine the legal validity and evidentiary status of electronic transactions, but do not independently establish moral responsibility for AI decisions.

F. Constitutional principles

The Constitution of the Kingdom of Thailand B.E. 2560 (2017) provides important background principles concerning equality, human dignity, rights, and freedoms.

These values may inform the interpretation of applicable legislation and the assessment of unlawful conduct.

However, constitutional rights should not automatically be treated as independent private damages claims without examining the relevant legal basis.

3. Meaning and scope of moral AI liability

Moral AI liability involves the relationship between technological conduct and legally protected human interests.

It is useful to distinguish five dimensions.

A. Ethical responsibility

Ethical responsibility concerns whether an AI system has been designed and operated according to acceptable standards of fairness, transparency, safety, and accountability.

Examples:

AI systems that discriminate against women in recruitment.

Automated lending systems that disadvantage particular communities.

AI systems that manipulate vulnerable consumers.

Facial recognition systems that incorrectly identify individuals.

Ethical misconduct does not automatically establish civil liability. The conduct must also satisfy the requirements of an applicable legal cause of action.

B. Human dignity and autonomy

AI systems can interfere with human autonomy when they make consequential decisions without meaningful safeguards.

Examples include:

Automated rejection of job applicants.

AI-generated medical treatment recommendations.

Unauthorised biometric monitoring.

Automated assessments of individual behaviour.

Manipulative psychological profiling.

Where these practices unlawfully interfere with legally protected rights, Section 420 of the CCC and applicable data-protection legislation may provide remedies.

C. Algorithmic discrimination

Moral AI liability may arise when automated systems produce discriminatory outcomes.

Relevant situations include:

Gender-based recruitment discrimination.

Age-based financial exclusion.

Disability-related discrimination.

Ethnic or social profiling.

Unequal access to healthcare.

Discriminatory educational assessments.

The claimant must identify the applicable protected right or statutory obligation and establish the necessary elements of liability.

D. Moral responsibility of developers

AI developers may face legal responsibility where they:

Ignore known system vulnerabilities.

Use unlawfully obtained training data.

Fail to conduct reasonable testing.

Misrepresent the reliability of their products.

Ignore foreseeable discriminatory outcomes.

Fail to correct known defects.

The developer's responsibility depends on its contractual role, control over the system, knowledge, conduct, and the applicable legal duties.

E. Institutional responsibility

A hospital, bank, university, employer, or government contractor cannot necessarily transfer its legal obligations to an AI supplier.

Where an institution makes consequential decisions using AI, its own conduct and legal duties remain relevant.

4. Six or more important Thai judicial decisions relevant to moral AI liability

Analogous precedents, not AI-specific judgments

Thailand has not established a substantial body of reported Supreme Court decisions specifically concerning civil liability for unethical or morally harmful AI decisions. Published Thai PDPA litigation is also limited.

The following judicial authorities are therefore useful for developing arguments concerning privacy, dignity, negligence, reputational harm, digital platforms, and accountability. They should not be presented as judgments directly deciding AI liability.

Case 1. Supreme Court Judgment No. 4893/2558 (2015)

Subject: Privacy, dignity, and unlawful publication

Facts and background:

The dispute concerned the publication of intimate photographs of private individuals without their consent.

The affected individuals argued that the publication violated their privacy and caused personal and reputational injury.

Legal issues:

Whether privacy constitutes a legally protected right.

Whether publication of intimate material without consent is unlawful.

Whether the conduct constitutes a wrongful act under Section 420.

Whether freedom of expression can justify interference with private life.

Legal significance:

The decision is relevant to the protection of personal privacy as a legally recognised interest.

Application to moral AI liability:

This decision may provide an analogous foundation for claims involving:

AI-generated intimate images.

Deepfake sexual imagery.

Unauthorised facial recognition.

AI-generated disclosure of private information.

Automated exposure of confidential personal data.

An AI operator that knowingly or negligently facilitates unlawful disclosure may face civil liability where the relevant legal requirements are established.

Case 2. Supreme Court Judgment No. 15067/2557 (2014)

Subject: Medical negligence and inaccurate information

Facts:

A healthcare institution and medical professional were involved in providing an inaccurate medical examination report.

The incorrect reporting delayed the identification of the claimant's serious medical condition.

Legal issues:

Whether inaccurate medical information constitutes negligence.

Whether professional responsibility arises from incorrect reporting.

Whether the resulting harm was sufficiently connected to the defendants' conduct.

Whether compensation was available.

Legal significance:

The decision demonstrates that inaccurate information supplied through a professional service can create civil responsibility when it causes foreseeable harm.

Application to moral AI liability:

It is particularly relevant to AI healthcare systems.

For example, if a hospital relies on an AI diagnostic tool that produces an inaccurate result, the hospital and relevant professionals may face liability depending on their duties, supervision, and negligence.

The developer may also face a separate claim if defective design or inadequate warnings caused the injury.

Case 3. Supreme Court Judgment No. 1808/2561 (2018)

Subject: Concealment of material information and wrongful conduct

Facts:

A property developer sold property without adequately disclosing important information concerning expropriation affecting part of the property.

The purchasers suffered financial consequences.

Legal issues:

Whether concealing important information can constitute a wrongful act.

Whether the conduct violates the purchaser's legally protected interests.

Whether compensation is recoverable under Section 420.

Legal significance:

The decision illustrates the legal importance of material information and the consequences of withholding information that affects another person's interests.

Application to moral AI liability:

It can be applied by analogy to AI services where organisations:

Conceal important algorithmic limitations.

Misrepresent AI accuracy.

Hide known discriminatory outcomes.

Fail to disclose material risks.

Mislead consumers about automated decision-making.

The case does not establish a general legal duty to disclose every detail of an algorithm.

Case 4. Chiranuch Premchaiporn case – Supreme Court Judgment No. 1167/2553

Subject: Online intermediary responsibility and digital content

Facts:

The case concerned the Prachatai online discussion platform and allegedly unlawful third-party comments published through its forum.

The proceedings involved the responsibility of the platform's director under Section 15 of the Computer Crime Act B.E. 2550.

Legal issues:

Responsibility of online service providers.

Management of user-generated content.

The relationship between platform operators and third-party publications.

Statutory responsibilities concerning unlawful information.

Legal significance:

The case is important for understanding the distinction between content creators and intermediaries.

Application to moral AI liability:

Its principles are relevant by analogy to:

AI-generated content platforms.

Automated moderation systems.

Generative AI services.

AI-powered social networks.

Platforms that distribute harmful synthetic content.

The case should not be interpreted as establishing automatic civil liability for all AI-generated content.

Case 5. Natural Fruit Co. Ltd. v. Andy Hall

Prakanong Provincial Court, Black Case No. 1150/2557, Red Case No. 843/2559; Appeal Court decision in 2019

Subject: Defamation, reputational injury, and responsibility for information

Facts:

Researcher Andy Hall participated in research concerning working conditions in Thailand's food-processing industry.

Statements and research findings were communicated to journalists and published online.

Natural Fruit pursued civil proceedings alleging reputational damage.

Legal issues:

Whether inaccurate factual statements damage another person's reputation.

Whether the defendant should have known that the statements were untrue.

Whether the publication was protected by a legitimate interest or fair comment.

Whether compensation was available under Section 423 of the CCC.

Legal significance:

The civil proceedings demonstrate that persons communicating information that harms another's reputation may face compensation claims if the statutory requirements are satisfied.

The initial civil judgment ordered THB 10 million in damages, and the appeal judgment upheld that outcome.

Application to moral AI liability:

This decision is relevant to:

AI-generated defamatory statements.

False allegations generated by chatbots.

Automated reputational scoring.

AI-generated misinformation about individuals.

Synthetic content damaging commercial reputation.

The case also illustrates why the truthfulness, verification, and context of information matter when evaluating automated content.

Case 6. Supreme Court Judgment No. 2738/2554 (2011)

Subject: Banking negligence and financial responsibility

Facts:

The dispute involved forged cheques and the bank's handling of a financial transaction.

A bank employee certified a cheque bearing a forged signature and company seal, resulting in a financial loss.

Legal issues:

Whether the bank employee acted negligently.

Whether the bank was responsible for the employee's conduct.

Whether the customer also contributed to the loss.

How responsibility should be apportioned.

Legal significance:

The judgment illustrates that financial institutions may be responsible for negligent conduct in processing financial transactions, while the claimant's own negligence may also be considered.

Application to moral AI liability:

This reasoning is relevant to AI-powered financial decisions, including:

Automated credit rejection.

AI-based fraud detection.

Incorrect account suspension.

Automated transaction approval.

Unfair financial-risk scoring.

An institution cannot automatically avoid responsibility by claiming that its algorithm made the decision.

Case 7. Supreme Court Judgment No. 6010/2564 (2021)

Subject: Financial misconduct, participation, and responsibility

Facts:

The case involved the diversion of company funds through bank accounts belonging to different individuals.

The Court considered the responsibility of persons who received funds transferred as part of the scheme.

Legal issues:

Whether the defendants participated in the wrongful financial arrangement.

Whether receipt of money established responsibility.

Whether the available evidence demonstrated involvement.

Whether repayment of the diverted funds was required.

Legal significance:

The decision illustrates the importance of establishing individual conduct and participation before attributing legal responsibility.

Application to moral AI liability:

It offers an analogous approach to determining the responsibilities of multiple parties involved in an AI system.

For example, liability should be examined separately for:

The AI developer.

The organisation deploying the AI.

The institution using the output.

The individual approving the decision.

The third-party data provider.

Responsibility should not be attributed merely because a party is associated with the technology.

5. Comparative case-law analysis

These decisions provide different legal analogies. None establishes an independent doctrine of moral responsibility for artificial intelligence.

6. Civil liability of different AI stakeholders

Moral AI litigation requires identifying the person or organisation that had the relevant legal responsibility.

A. Developer liability

An AI developer may face claims under Section 420 where negligent development causes unlawful harm.

Potential allegations include:

Failure to remove known discriminatory biases.

Inadequate training-data verification.

Failure to conduct appropriate safety testing.

Failure to address known system vulnerabilities.

Misrepresentation of system accuracy.

B. Operator liability

An organisation deploying an AI system may face responsibility where it fails to assess risks or exercise appropriate supervision.

For example, a bank that automatically rejects customers using an inadequately tested algorithm may face claims if its conduct violates an applicable legal duty and causes compensable harm.

C. Employer liability

Section 425 may make employers jointly liable for wrongful acts committed by employees in the course of employment.

This is relevant where employees improperly use AI systems to evaluate customers, employees, patients, or other individuals.

D. Data controller liability

Under PDPA Sections 77 and 78, a data controller may face civil liability and potentially punitive damages where the statutory requirements are satisfied.

This is particularly relevant to AI systems that use sensitive personal data, behavioural profiling, biometric identification, and automated predictions.

7. AI ethics and judicial governance in Thailand

An important development is Thailand's movement toward formalising responsible AI use.

A. Thailand's AI ethics guidelines

Thailand has developed AI ethics guidance addressing principles such as:

Fairness.

Transparency.

Accountability.

Human-centred development.

Privacy protection.

Safety and security.

Responsible innovation.

These guidelines provide important standards for evaluating responsible AI deployment, although ethical guidelines do not automatically create independent civil causes of action.

B. Supreme Court recommendations of 19 September 2025

The President of the Supreme Court issued recommendations concerning AI technology in judicial case operations.

The recommendations address the use of AI within the justice system and emphasise fairness, impartiality, judicial independence, and responsible use.

They are important because judicial decision-making must remain accountable to human legal institutions.

Oxford Institute of Technology and Justice+1

C. Civil Court Regulations B.E. 2568 (2025)

Important regulatory development

The Civil Court introduced regulations governing the use of AI when preparing pleadings and other documents submitted to the court.

The framework requires:

Disclosure of AI-generated portions of court documents.

Identification of content generated using AI.

Verification of factual statements and legal provisions.

Certification by the submitting party or lawyer.

Continued human responsibility for the accuracy of submitted material.

These requirements are particularly significant for moral AI liability because they establish that a person cannot simply blame an AI system for inaccurate material submitted to a court.

The regulations also contemplate sanctions for non-compliance, including potential contempt-of-court consequences under the relevant procedural provisions.

Oxford Institute of Technology and Justice+1

8. Important hypothetical litigation scenarios

The following examples illustrate how moral AI liability could arise in Thailand.

Scenario 1: AI recruitment discrimination

A Thai company deploys an AI recruitment system that systematically rejects qualified female applicants.

Potential legal issues:

Unlawful discrimination.

Negligent algorithmic deployment.

Employment-related statutory protections.

Civil liability under Section 420, where its elements are established.

Potential defendants include the employer and, depending on the circumstances, the AI developer.

Scenario 2: AI healthcare discrimination

A hospital deploys an AI diagnostic system that consistently produces less accurate recommendations for a particular patient group.

Potential legal issues:

Medical negligence.

Failure to supervise automated recommendations.

Inaccurate medical information.

Personal injury compensation.

The hospital and responsible healthcare professionals may face liability depending on their duties and conduct.

Scenario 3: AI-generated reputational injury

An AI chatbot generates false allegations about an individual and distributes them publicly.

Potential legal issues:

Defamation.

Section 423 of the CCC.

Section 420 where applicable.

Platform responsibility.

Electronic evidence and attribution.

Scenario 4: AI emotional manipulation

A commercial application uses AI to exploit users' emotional vulnerabilities and persuade them to make unnecessary purchases.

Potential legal issues:

Unfair commercial practices.

Misleading representations.

Consumer protection.

Unlawful personal-data processing.

Scenario 5: AI financial exclusion

An AI-powered lending platform rejects applicants through an unfair and inaccurate scoring mechanism.

Potential legal issues:

Negligence.

Inaccurate personal-data processing.

Consumer protection.

Contractual obligations.

Discriminatory conduct, where prohibited by applicable law.

Scenario 6: AI deepfake exploitation

An individual uses generative AI to create and distribute intimate images of another person without consent.

Potential legal issues:

Privacy violations.

Section 420 tort liability.

PDPA violations where applicable.

Criminal liability under relevant legislation.

Compensation for legally recognised damage.

9. Evidence required in moral AI liability litigation

Successful AI-related civil litigation may depend on technical and documentary evidence.

The claimant generally bears the burden of proving the elements of the relevant civil claim, subject to applicable statutory exceptions and evidentiary rules.

A particular difficulty is proving causation where an AI system operates as a complex or opaque decision-making mechanism.

10. Remedies available to victims

Victims of morally harmful AI practices may seek different remedies depending on the nature of the violation.

1. Compensation under Section 420

Compensation for legally recognised injury caused by unlawful intentional or negligent conduct.

2. Compensation under PDPA Section 77

Compensation for damage resulting from violations of applicable personal-data protection requirements.

3. Punitive damages under PDPA Section 78

Where the statutory requirements are met, the court may award punitive damages subject to the statutory maximum of twice the actual damages.

4. Contractual remedies

A victim may seek contractual compensation, termination, or other appropriate relief where an AI service provider has breached contractual obligations.

5. Consumer remedies

Consumers may pursue remedies under applicable consumer-protection legislation.

6. Injunctive relief

Where the relevant legal requirements are satisfied, a claimant may seek appropriate court orders to prevent continuing unlawful conduct.

7. Class-action litigation

Where multiple persons suffer similar harm, collective proceedings may be available under Thailand's class-action framework.

11. Major challenges in Thai moral AI liability litigation

Several unresolved issues are likely to shape future disputes.

AI legal personality: Whether AI should ever receive an independent legal status.

Algorithmic causation: How courts should establish the connection between automated decisions and individual harm.

Moral damages: How non-economic harm involving dignity, autonomy, and psychological injury should be assessed.

Developer responsibility: How liability should be divided between developers and operators.

Algorithmic transparency: How claimants can obtain sufficient evidence about automated systems.

Cross-border liability: How Thai courts should address foreign AI service providers.

Human oversight: What degree of supervision is reasonable for high-impact AI applications.

Ethical standards: Whether AI ethics guidance should influence the assessment of negligence and professional standards.

12. Conclusion

Moral AI liability litigation in Thailand is an emerging area of civil law that connects traditional tort principles with modern technology, privacy, consumer protection, and human dignity.

The Civil and Commercial Code, especially Sections 420, 421, 425, 437, 438, and 448, provides important foundations for determining responsibility and compensation.

The PDPA adds specific protections against unlawful personal-data processing, while consumer-protection legislation addresses unfair AI-enabled commercial conduct.

The judicial authorities discussed above provide useful analogies concerning privacy, negligence, reputational injury, financial responsibility, and intermediary conduct. They do not establish six AI-specific judicial precedents.

Thailand's 2025 judicial AI recommendations and Civil Court AI-document regulations are particularly important developments because they reinforce the principle that human institutions and professionals remain accountable when using AI.

The central principle is that artificial intelligence may assist human decision-making, but its use does not automatically transfer or eliminate the legal responsibility of the people and organisations controlling, deploying, or relying upon it.

Research note: This is an academic legal overview, not a claim that Thailand has already recognised a standalone tort of moral AI misconduct. The precise outcome of litigation depends on the statutory cause of action, evidence, and circumstances of each case.