You are using an outdated browser and your browsing experience will not be optimal. Please update to the latest version of Microsoft Edge, Google Chrome or Mozilla Firefox. Install Microsoft Edge

December 3, 2025

Thailand Civil Court Issues Rules on AI-Generated Court Filings

Subscribe to Updates

Thailand’s Civil Court has issued a regulation targeting the use of artificial intelligence (AI) in the preparation of pleadings and other documents submitted to the court. Effective November 17, 2025, the regulation aligns with September 2025 guidance from the president of the Supreme Court, and aims to safeguard accuracy, transparency, and public confidence in civil adjudication.

The regulation applies to all parties submitting pleadings or any documents to the Civil Court that are prepared using AI tools or contain AI-generated content. It subjects AI used for these purposes to strict requirements on verification, disclosure, and accountability.

Core Obligations

The regulation imposes four principal obligations:

  • Lawyers who use AI remain subject to duties of honesty, responsibility to the court, professional standards, and legal ethics, including the duty to assess the appropriateness of the AI tool for the work.
  • Parties and lawyers must verify the accuracy and completeness of all facts, legal provisions, and citations in AI-generated content before submission.
  • Parties and lawyers must disclose to the court any AI-generated content by clearly marking the beginning and end of the AI-generated portion with prescribed statements (see below). Additionally, a certification confirming the use of AI must be provided at the end of the pleading or document, stating that AI was used for certain portions and that the party has reviewed and certifies the accuracy of factual and legal content.
  • Parties and lawyers bear the same full legal and ethical responsibility for AI-generated content as they do for personally authored documents; they cannot evade responsibility or avoid liability by citing AI-related errors. Likewise, parties must ensure that any AI-generated content is truthful, accurate, and unbiased.

Prescribed Disclosure Language

Each instance of AI-generated content must be preceded by the statement “[The following content was prepared using artificial intelligence]” and must end with “[End of content prepared using artificial intelligence].”

In addition, any filing that includes AI-generated content must include the following certification at the end of the document: “Certain portions of this [type of document] were prepared using artificial intelligence. I have reviewed and certify the accuracy of the facts and the law prior to submission to the court.”

Enforcement and Compliance Considerations

Noncompliance may be reported by a court officer or judge to superiors for further orders and may constitute disorderly conduct within court precincts under the Civil Procedure Code. In practical terms, counsel should anticipate potential adverse procedural measures, reputational consequences, and other court-ordered remedies for violations.

It is important for lawyers to implement AI-use protocols that align with the regulation’s requirements. This includes selecting AI tools solely as aids to human drafting, ensuring that every AI-generated output undergoes thorough human review for accuracy of facts, laws, and citations, inserting the exact disclosure statements around each AI-generated section, and adding the mandated end-of-document certification.

Lawyers must also apply litigation quality controls—such as primary-source verification, adversarial checks for hallucinations or bias, and chronological and citation audits—before filing. For mixed-origin documents, lawyers must maintain internal versions that identify which sections are AI-assisted to avoid disclosure gaps.

Finally, firms must update professional responsibility guidance and training to reinforce that lawyers remain wholly accountable for AI-assisted content.

Key Takeaways

The Civil Court’s regulation recognizes the legitimate role of AI in modern litigation while requiring human oversight, explicit transparency, and uncompromised professional responsibility. Legal practitioners and parties may continue leveraging AI’s efficiencies, provided they embed rigorous review processes and the mandated disclosures into the filing workflows.

The regulation is also the latest in a broader effort by the Thai courts to address the implications of AI use across the judiciary. Aside from the regulation, the president of the Supreme Court also issued guidance in September 2025 promoting ethical and cautious use of AI technologies in court operations.

Related Professionals

Industries

Technology

Practices

Dispute Resolution and Litigation

Location

RELATED INSIGHTS​ 

December 26, 2025
The Bank of Thailand (BOT) has released the Guidelines for Digital Fraud Management, which took effect on December 17, 2025, incorporating certain amendments to the draft guidelines issued in March 2025. These official guidelines aim for end-to-end digital fraud prevention, with a particular focus on mule accounts, to enhance trust and security in Thailand’s financial system. The guidelines apply to “financial service providers,” including: Financial institutions and special financial institutions under the Financial Institution Business Act; and Operators of Inter-institutional Fund Transfer System e-money services and e-fund transfer services under the Payment Systems Act. Besides commercial banks and e-money operators that offer fund-transfer services, other providers may adopt requirements based on risk proportionality and baseline standards set out in the guidelines (for instance, an e-money operator that does not offer e-fund transfer services could consider implementing a fraud monitoring and detection system according to the risk level of its service). The guidelines establish the following key requirements: Policy and oversight. Directors and senior executives of financial service providers must adopt appropriate “end-to-end” fraud management policies and KPIs to manage digital fraud, covering prevention, monitoring, detection, management, resolution, and support for affected customers. The fraud management policy must be regularly reviewed, and whenever there is a situation or change that significantly affects the efficiency of the fraud management. Any significant update to the policy must first be approved by the board of the financial service provider. The BOT also encourages providers to collaborate in establishing industry standards aligned with applicable laws and regulations to ensure consistency and best practices across the sector. Fraud management processes. Financial service providers must establish a clear framework for managing digital fraud throughout the customer lifecycle—from customer onboarding to service termination—covering at least the following processes: Know your customer (KYC) and customer due diligence (CDD):
Read More
December 15, 2025
Thailand is taking steps to energize its startup scene by drafting the Startup Promotion Law. This draft law aims to remove obstacles, open new funding opportunities, and provide coordinated government support. The goal is to make it easier for Thailand-based startups to grow and compete on a global stage. Why Is This Law Needed? For many years, Thai startups have operated under traditional company law frameworks that were not designed with high-growth businesses or with fundraising opportunities in mind. Restrictions on issuing bonds, offering shares to outside investors, and repurchasing shares for employee incentive programs made it challenging for emerging companies to access capital and accelerate their growth. The draft Startup Promotion Act seeks to remove these obstacles and foster a more competitive, entrepreneur-friendly environment in Thailand. Who’s in Charge? Two main organizations will oversee the startup ecosystem: Startup Promotion Committee: This group, to be appointed by the National Science, Research, and Innovation Policy Council, will set national strategies, policies, and budget; design promotional campaign and incentives; and propose further legislative amendments to promote startups. National Innovation Agency (NIA): Under the draft act, the NIA will be the main contact for startups and will serve as the secretariat office of the Startup Promotion Committee, coordinating data, advising startups, maintaining the public registry, and providing funding and investment (grants, repayable grants, loans, and equity) under committee criteria and, where applicable, cabinet approval. What Startups Are Eligible for Benefits? To be officially recognized and access benefits, a company must: Be a private limited company less than 10 years old at the time of application. Existing companies that already exceed the 10-year threshold may still apply for startup statues within one year of the law’s enactment, as long as they otherwise still qualify for the new regime. Have average annual revenue not
Read More
December 12, 2025
Tilleke & Gibbins has updated the Cambodia, Myanmar, Thailand, and Vietnam chapters in Multilaw’s Global Data Protection Guide, which collects expert advice from Multilaw member firms in 90 jurisdictions around the world (including a Laos chapter, which is also authored by Tilleke & Gibbins). The guide provides answers to key issues concerning the fast-developing data protection and privacy laws around the world, and helps data protection officers and in-house counsel understand how the regulatory regime for data protection can affect their organizations in various jurisdictions. Each section of the guide identifies the main laws that govern data protection in that jurisdiction, and gives a detailed overview of the legal principles in place as well as the enforcement authorities responsible for overseeing compliance. The guide also covers issues related to data subject rights, data protection officers, impact assessments, data breach notification requirements, and cross border data transfers. The use of personal data in marketing is also considered, with specific information on electronic marketing rules, cookies, and marketing to businesses and consumers. Multilaw, of which Tilleke & Gibbins is a longtime member, is a global network of carefully selected independent law firms able to provide expert legal advice in complex environments around the globe. The full guide is available for free on the Multilaw website.
Read More
December 11, 2025
On December 10, 2025, the National Assembly of Vietnam passed a new Cybersecurity Law, which will take effect on July 1, 2026. The new Cybersecurity Law was developed based on the consolidation of the 2018 Cybersecurity Law and the 2015 Law on Network Information Security. While the final approved version of the new Cybersecurity Law has not yet been published, according to official reports, the following notable requirements are confirmed to be included: The new Cybersecurity Law dedicates a specific article to prohibited acts related to cybersecurity, under which it strictly prohibits posting or disseminating information online that propagandizes against the Socialist Republic of Vietnam. The law also prohibits, among other things, (i) the appropriation, trading, seizure, or intentional disclosure of information classified as state secrets, work secrets, business secrets, personal secrets, family secrets, and private life; (ii) intentionally eavesdropping, recording, or filming online conversations without authorization; and (iii) the use of artificial intelligence (AI) or new technologies to conduct prohibited acts. The Ministry of Public Security (MPS) has the authority to require enterprises providing telecommunications, internet, and online services, as well as system administrators, to remove information violating cybersecurity laws from systems under their management. The MPS is also assigned responsibility for ensuring information security in cyberspace and data security, establishing mechanisms for IP address identity management, verifying digital account registration information, and issuing warnings and sharing information on cybersecurity threats. Information systems are classified into five levels (similar to the 2015 Law on Network Information Security) based on the degree of harm to national security and social order if an incident occurs. The MPS is the lead agency assisting the government in state management of cybersecurity. The Ministry of National Defense is responsible for managing military information systems, and the Government Cipher Committee manages cryptographic and cipher
Read More