REPORT OF RESULTS Workshop on Skills and Experiences of Judges in Applying Artificial Intelligence in Resolving Civil Cases

REPORT OF RESULTS
Workshop on Skills and Experiences of Judges in Applying Artificial Intelligence in Resolving Civil Cases

 
In accordance with the Cooperation Agreement between the Judicial Academy and the National School of Judges of the Republic of France (RFR) signed in 2012, and with the approval of the Ministry of Justice, the Judicial Academy (Academy) organized a workshop on "Skills and Experiences of Judges in Applying Artificial Intelligence in Resolving Civil Cases" on the afternoons of November 17th and 18th, 2025. The Academy would like to report the results of the workshop as follows:
I. GENERAL INFORMATION ABOUT THE WORKSHOP
1. Workshop Title: Skills and Experiences of Judges in Applying Artificial Intelligence in Resolving Civil Cases.
2. Workshop Objectives
- To raise awareness and understanding among the Academy's lecturers and students regarding the role, potential, and challenges of applying AI in the resolution of civil cases.
- To introduce and disseminate international experience, especially that of the Republic of France, regarding the legal framework governing AI and the skills and experience of judges in using AI in the stages of resolving civil cases.
- To enhance the capacity of the Academy's lecturers in the necessary knowledge and skills when applying AI in judicial activities, including skills in using AI tools, skills in assessing and controlling risks, and ensuring professional ethics.
3. Time, Location, and Format of the Workshop The workshop will be held over two afternoons on November 17-18, 2025, online via Zoom, with two connection points at the National School of Judges of the Republic of France and the headquarters of the Judicial Academy in Hanoi, simultaneously connecting online with experts in the Republic of France and delegates in Vietnam.
4. Content, Program, and Communication Activities of the Workshop
4.1 The workshop includes 04 working sessions according to the program as follows:
- Afternoon session on November 17, 2025, on 02 topics:
+ Working session 1: Using AI for judicial efficiency: For the Supreme Court of the Republic of France.
+ Working session 2: Skills and experience of judges when using AI in resolving civil cases in Vietnam.
- Afternoon session on November 18, 2025, on 02 topics:
+ Working session 3: Ensuring transparency and professional ethics of judges when using AI in resolving civil cases in the Republic of France.
+ Working session 4: Enhancing the capacity of judges in applying AI in the judicial activities of Vietnam.
4.2. Conference Information and Communication: The conference activities were reported on the Ministry of Justice's electronic portal and the Vietnam Law Newspaper.
5. Chairperson: Assoc. Prof. Dr. Nguyen Minh Hang, Deputy Director of the Judicial Academy.
6. Organizing Agency/Unit: The conference was chaired by the Judicial Academy, in coordination with the National School of Judges of the Republic of France.
7. Speaker Information:
- The conference had two Vietnamese speakers: Judge Huynh Ngoc Tuan, People's Court of Region 1, Ho Chi Minh City, and Ms. Luu Ngoc Lien, Lecturer in the Department of General Training for Judges, Prosecutors, and Lawyers, Judicial Academy.
 The conference had two foreign speakers: 1/ Ms. Ferreira Laisa, Mylene, Researcher at the Research Department, National School of Judges of the Republic of France; 2/ Mr. Allain Matthieu, Marie, Judge of the Supreme Court of the Republic of France. 8. Participants, number, and composition of Vietnamese delegates and delegates with foreign nationality
8.1. The Vietnamese side has 141 delegates participating, including:  Assoc. Prof. Dr. Nguyen Minh Hang, Deputy Director of the Judicial Academy.
 Delegates from the Department of International Cooperation, Central Vietnam Law College, Southern Vietnam Law College, Department of Internal Political Security, Ministry of Public Security, and reporters from Vietnam Law Newspaper.
 Leaders and lecturers of the Faculty of Training for Judges, Prosecutors, and Lawyers; lecturers of the Faculty of Lawyer Training and the Department of Training for Lawyers, Judges, and Prosecutors at the Ho Chi Minh City campus; 3 visiting lecturers participating in teaching the lawyer training program and the general training program for judges, prosecutors, and lawyers in Hanoi and Ho Chi Minh City;
8.1. The Head of Department and 2 specialists from the Department of Administration, Management and External Relations; Editor of the Judicial Academy's online portal; trainees participating in joint training courses for judges, prosecutors, and lawyers, and professional lawyer training courses in Hanoi and Ho Chi Minh City.
8.2. The French side has 5 delegates, including:
+ Participants from the French Embassy in Vietnam: Mr. Romain Busuttil, Deputy Counselor for Cooperation and Cultural Activities, French Embassy in Vietnam, and Ms. Priscille de Cambourg, Officer in charge of judicial cooperation, French Embassy in Vietnam.
+ Ms. Emmanuelle LAUDIC-BARON, Head of the International Cooperation Department, National School of Judges of the Republic of France, and 2 speakers from the Republic of France.
9. Funding
- The French Embassy in Vietnam and the National School of Judges of the Republic of France will cover the costs of hiring interpreters and experts from the Republic of France.
- The Vietnam Judicial Academy covered the costs of renting translation equipment, Zoom software, hiring Vietnamese experts, expenses for delegates attending, and other costs related to organizing the workshop in Vietnam.
II. WORKSHOP RESULTS
1. Workshop Results After the opening ceremony, the workshop worked on the following topics:
• Working Topic 1 on the use of AI for judicial efficiency: Presented by Mr. Allain Matthieu, Judge of the Supreme Court of the Republic of France.
In his presentation, Mr. Allain Matthieu stated that the Court of Cassation is one of the world's pioneering judicial bodies in applying AI in a controlled and responsible manner, and maintains a close cooperative relationship with the French National School of Judges (ENM) in research, implementation, and judge training.
Several notable AI projects have been successfully implemented at the Court of Cassation: First, an algorithm that automatically anonymizes entire judgments before public release – a mandatory requirement under the 2016 Law on Open Judicial Data. Next is the “organizational memory amplification” project, which allows different departments (civil, commercial, social, criminal, etc.) to instantly access and reuse knowledge from millions of old judgments, while also providing open data sources for other judicial bodies in France. The project supports data scientists in processing case files automatically five times faster than humans, with AI engineers trained to understand the specialized work of each court, thereby providing accurate and efficient support. In particular, the “Di Vec Gioăng” project (automated case law analysis) is the most ambitious project, having inputted millions of judgments from first instance courts, appellate courts and the Court of Cassation, achieving a very high accuracy rate in analysis. Although only initial results are available, it has been identified as a strategic priority project of the Court of Cassation in the 2024-2028 period.
In 2025, the Chief Justice of the Court of Cassation established an inter-agency working group on artificial intelligence with diverse members: judges, rapporteurs, lawyers, assistants, technology engineers, representatives of specialized departments and the judges' union. The group worked for one year to accurately statistically assess the actual needs of judges, build a “Profile of AI Users” in the judicial system and propose 05 mandatory criteria to evaluate all new AI projects: (1) AI only replaces highly repetitive tasks, absolutely does not replace human judgment; (2) The application of AI must not reduce the quality of trials and must not create additional risks to the quality of work; (3) Simultaneously meet legal, technical, financial, energy consumption and environmental protection requirements (especially note that large graphics processing models consume a lot of electricity and require expensive cooling systems); (4) Ensure data sovereignty and security – prioritize storage within the national territory, minimizing the transfer of data to foreign servers; (5) Absolutely respect the fundamental rights of citizens, with the top priority being the right to be tried by a human judge, the right to access transparent justice, without discrimination and avoiding the risk of bias due to non-representative training datasets.
Since 2016, the European Data Protection Regulation (GDPR) and the EU AI Act 2024 have absolutely prohibited fully automated judicial decision-making – a regulation that applies directly in the Republic of France. The ultimate responsibility always rests with the human judge and cannot be transferred to AI in any way.
Regarding the types of AI models used, the Court of Cassation employs a variety:
- Open-source models (internally regulated).
- Large-scale commercial models (used only in isolated environments).
- Specialized models developed by French universities and businesses specifically for the judiciary. Mr. Matthieu emphasized that while commercial models from large corporations offer superior processing power, judicial bodies often have to use closed models, which are less efficient but safer in terms of data security and sovereignty.
Currently, AI is used daily by judges of the Court of Cassation to summarize case files, search for and suggest precedents, assist in drafting judgments or rapporteur opinions, suggest applicable laws, and detect inconsistencies in reasoning. However, all uses exclude the possibility of automated judgments; judges are required to control, modify, and be 100% responsible for the final content. If a judge copies verbatim a text generated by AI without understanding, controlling, or explaining it, it will be considered a serious violation of professional ethics.
Regarding AI governance: Priority should be given to projects that deliver quick and low-cost benefits; effective research should be combined between technical departments and judges; and awareness of AI among judges and legal experts should be raised. The French Ministry of Justice is also strongly implementing digital transformation and AI throughout its entire system. Artificial intelligence is an extremely powerful tool that helps increase work efficiency without requiring excessive investment, but the red line that must never be crossed is replacing human judges in the process of adjudication. The right to be judged by a human judge, fairly and transparently, must always be prioritized.
• Working session 2 on the skills and experience of judges using AI in resolving civil cases in Vietnam was presented by Judge Huynh Ngoc Tuan, People's Court of Region 1, Ho Chi Minh City.
Judge Huynh Ngoc Tuan stated: Since 2023, the People's Court of Ho Chi Minh City has pioneered the deployment of artificial intelligence (AI) to support almost the entire process of resolving civil cases, from the acceptance stage to the issuance of judgments.
- During the case processing phase, AI automatically categorizes the lawsuit, extracts information on the parties involved, identifies the nature of the dispute, checks the statute of limitations and jurisdiction, and automatically suggests judges and clerks based on a random assignment principle, while also immediately alerting any errors.
- During the trial preparation phase, AI summarizes thousands of pages of case files in minutes, searches for and suggests precedents and similar legally binding judgments from the national database, detects contradictions in the arguments of the parties, and suggests applicable laws, resolutions, and circulars.
- During the judgment drafting phase, AI assists in drafting the reasoning and decision sections, offers multiple solutions for the judge to choose from, and automatically checks for spelling errors, formatting, and legal citations. - Artificial intelligence is merely a supporting tool, helping judges work faster, more comprehensively, and in a more multifaceted way, but it absolutely cannot replace humans in making final judgments. Judges must still review, revise, select, and bear full legal responsibility for the signed judgments.
- Limitations in the current use of artificial intelligence in courts: The quality of input data is inconsistent, the accuracy of AI has not reached 100%, and there are still cases of incorrect precedents or laws being proposed; a segment of older judges are still hesitant about the technology, especially since the legal framework regulating the use of AI in litigation is still incomplete, particularly regarding the legal value of electronic documents, electronic signatures, and the responsibility of judges when using AI.
 
- The National Assembly and the Government need to quickly finalize the legal framework on digital transformation in civil litigation; the Supreme People's Court needs to issue specific regulations on the use of AI in trials; Simultaneously, incorporate AI training content into the basic and continuing education programs for judges and clerks; continue to learn from the experience of the Republic of France, especially in developing ethical rules and criteria for the use of AI by judges.
* Former judge and lawyer Nguyen Huyen Cuong posed questions to Vietnamese and French experts.
- The application of AI is a very complex issue. AI serves justice but should not weaken the core role of judges. What is the boundary?
- AI can assist in handling, summarizing, evaluating, and analyzing shortcomings in litigation; provide direction in judicial proceedings; so what is the impact of AI throughout the litigation process and evidence gathering?
- With the same set of documents, judges may have different perceptions and assessments at different levels and with different evaluation results. Judges have different bases for their assessments How can we determine which assessment is correct, and what are the general criteria for evaluating arguments?
- When assessing whether the collected documents are complete, what criteria are used to select evidence regarding procedural matters or content for making assessments?
- With a precedent, a judge can apply it based on complete data or when there are not enough details to match the precedent. What criteria can AI use to select the data to apply?
- In a case that may involve multiple legal relationships, how can AI assist in clearly identifying the issues that need to be resolved?
* Judge Huynh Ngoc Tuan shared:
AI is only a supporting tool. The judge, as the person directly handling the case, needs to determine the necessary documents and evidence in the file. The judge must communicate with the AI ​​to find out what the AI ​​needs to do, rather than asking the AI ​​to determine whether the documents are complete or not. If there is content that needs further clarification or alternative options, AI can provide guidance for further research and data collection. The more we interact with the AI, the better the results. Judges using AI must verify the information provided.
Regarding the application of precedents: AI can find similar judgments and precedents in terms of legal relationships. Judges can select suggested judgments and precedents provided by the AI ​​for reference in resolving cases.
* Former judge and lawyer Nguyen Thi Ha shared:
The research and decision-making process of judges affects the judge and those involved in the decision. Referencing AI can help judges make more comprehensive assessments during the research and final decision-making process. In practice, many judges, when facing difficulties, can consult with colleagues, ensuring the legal rights and interests of the parties involved.
* Lecturer Pham Nhu Hung, former Deputy Director of the Judicial Academy, asked:
- According to the experts' sharing, AI is currently being used at the Court of Cassation, the higher-level court. In my opinion, all levels of courts need to apply AI, so I would like the expert to share more about the application of AI in courts at all levels in the Republic of France. - AI should be encouraged to assist in drafting judgments and arguments. However, the hypothesis is that if an AI-drafted judgment is given to a judge but the judge does not edit it and directly uses that judgment, how is the judge's legal responsibility determined? Will the AI ​​be held accountable? AI can produce (relatively) accurate judgments; in the future, it cannot be asserted that AI cannot draft more complete judgments than humans.
- AI does not affect the professional code of ethics for judges. In the process of resolving cases, an open mindset is needed, and the use of AI is increasing. Therefore, specific regulations are needed in the code of professional ethics for judges. I have consulted the code of ethics for judges in the Republic of France. The Supreme People's Court in Vietnam is also considering revising its code of professional ethics for judges.
* Mr. Matthieu Allain commented:
AI is fully capable of drafting judgments, that's undeniable. However, AI cannot yet determine the quality of the document and the basis of the judgment. If humans can train it, AI can draft judgments like humans, using new technology to create a system that can process and divide the content into sections; but for generated AI, AI can completely learn how to draft judgments. However, using AI to completely replace humans is not yet possible because it involves many factors. Closed AI (often used for state agencies) has many limitations to develop focused on a specific task. The theory of AI emerged early on, and it took us 80 years for AI to become functional and applicable at this point in time. In the future, we will not have an AI powerful enough to think critically, consider multiple factors, and ensure that no two judgments are alike and that it incorporates human elements.
Currently, the code of professional ethics for lawyers includes provisions related to the application of artificial intelligence, and lawyers can also use AI. The application of AI by judges in their work is unavoidable, but they need to be subtle in making judgments or disclosing personal information, regarding confidentiality and professional ethics. Regulations in the code of professional ethics are necessary. Currently, there are no specific regulations on professional ethics in the judicial field related to the application of AI.
• Working topic 3 on ensuring transparency and professional ethics of judges when using AI in resolving civil cases in the Republic of France, presented by Ms. Ferreira Laisa, Researcher at the Research Department, National School of Judges of the Republic of France.
Ms. Ferreira Laisa stated: The core principle in professional practice is that the application of AI must ensure professional ethics. There's no denying the effectiveness of AI and its assistance to judges, as a tool to support them in repetitive tasks, analysis, and information processing.
First, it's necessary to clarify what artificial intelligence is. Over the past 20 years, there have been many different definitions in the EU. According to EU regulations in 2022, AI is considered software, but by 2024, the definition of AI became more complex, though fundamentally it remains software designed to have varying degrees of autonomy based on input data sources to perform analysis in a digital environment. This definition reflects future developments, differing from traditional software methods and representing an advanced form of software capable of self-learning.
Professional ethics: Ethics in general is a set of principles that guide general behavior and forms the framework for regulations concerning professional ethics. Lawmakers cite ethical issues not as legal regulations, and ethics can only be legal when ethics are incorporated into legal regulations. Professional ethics ensures legal legitimacy, institutionality, and the honesty of those participating in and conducting legal proceedings. Judges and prosecutors both demonstrate the role of the rule of law, shaping professional consciousness based on the ideal of integrity and institutionalizing it. The principle of judicial independence is constitutionally enshrined; technology should not be used to undermine this principle. The difference between predictive and forecasting methods.
The report was conducted by multiple units, reviewing needs according to each user group. Various units within the Ministry of Justice, such as secretarial offices, police, courts, and prisons, conducted surveys. Needs for document review and automatic document summarization emerged, with different agencies having their own specific requirements. The implementation faces legal challenges: the report clearly stated the need to protect fundamental human rights and privacy; data sovereignty: data control within the national scope requires human oversight, parameters, and oversight, but also requires judicial control to ensure that AI is operated and programmed by the judicial authorities and that data is controlled at a domestic data center. The increasing use of AI in various agencies is necessary; the existence of multiple artificial intelligence systems within the judicial system means that AI tools should not be used for decision-making, and humans should play a supervisory role. From the initial processing and comparison of documents to the allocation of files, the appropriateness of procedural documents, and the processing of legal information, precedents, and doctrines, AI assists in forming grounds and making judgments. Understanding the combined capabilities of various AI tools is crucial for addressing these issues effectively. The professional ethics of judges and the issue of enforcement control are governed by specific rules to ensure the independence of judges. Refusal of justice: judges' silence, reliance on AI due to judges' trust, AI lacking legitimacy; lack of competence in mastering AI technology due to slow adaptation; ambiguity when judges do not understand or cannot apply the results of AI, leading to an inability to explain the judgment. Judges must still ensure adherence to the adversarial principle and maintain control over their work.
The extent of AI interference in the work of judges currently needs to be determined: In 2016, the EU banned all decisions based on automated processing; all decisions must be made by humans. Hidden automation prohibition and legal liability: The European Union prohibits hidden automation, and while not yet enshrined in French law, it directly applies the responsibility for protecting personal data, including liability for data leaks by contractors and other stakeholders, with administrative penalties of up to 4% of the revenue of violating businesses. AI systems may also be subject to related penalties. There are no specific examples of agencies using AI without human intervention.
Challenges in ensuring data legality in France: Anonymizing sensitive information, the concept of anonymization of personal information, difficulties in achieving anonymity, and the need to comply with regulations regarding data disclosure Regulations on anonymity and handling of trial files to deliver judgments. Judges are responsible if personal information is disclosed. Anonymity in judgments depends on the stage at which the information is leaked and who is responsible at that stage. Information must be obtained from judges and prosecutors; data on previous judgments of a specific judge must not be used. If data from a previous judge's judgment is used, it will not be considered a legal basis. Responsibility will vary depending on the stage at which anonymity is required (from the initial input or at another stage).
• Working session 4 on enhancing the capacity of judges to apply AI in trial activities in Vietnam was presented by Ms. Luu Ngoc Lien, Lecturer, Faculty of General Training for Judges, Prosecutors, and Lawyers, Vietnam Academy of Justice.
Ms. Luu Ngoc Lien, M.A., stated: Regarding the issue of enhancing the application of AI in the judicial activities of judges in Vietnam: AI supports the administrative tasks of judges, especially those that are repetitive in nature. AI can optimize the process of analyzing and evaluating evidence, helping to reduce the workload of drafting documents, particularly in terms of form. From two perspectives, it can be seen that:
- Current legal regulations: In the Vietnamese court system, there are regulations on the application of information technology and digital transformation, especially in stages such as the delivery and receipt of lawsuits, documents, and evidence, requiring digital signatures to accept the applications. Electronic submissions still require printing and registration in the case register as if handling hard copies.
 
- How are judges currently applying AI? The application of AI in judicial activities in Vietnam is currently implemented synchronously and uniformly nationwide. Virtual Assistant: Judges and court clerks train the virtual assistant to provide information, present scenarios, and answer legal questions, providing keywords for use so that other units and subsequent users can search based on those keywords. Exchange and Expansion of Legal Cases: Data entry personnel upload information to the system for dissemination across the entire sector. Currently, the ability to analyze questions and input data sources does not meet the needs of those conducting the training. Without scenarios or keywords used, there are no results. The virtual assistant does not update to reflect new legal regulations. Human Resources Issues: The increasing volume of cases handled by the court system puts pressure on judges' professional work, thus limiting AI training. Speed, time, and the ability to access information will also take longer than with AI used by the general public. Information Exploitation and Security Issues: There are no established standards related to information exploitation and security. Proposal/Recommendation: Improve legal regulations and enhance the advantages of the 11 virtual assistants to achieve the expected effectiveness throughout the court system.
* Former judge and lawyer Nguyen Huyen Cuong raised the following questions: Ethics in the use of AI: In cases of ethical violations leading to risks of data bias, interpretation bias; risks of systemic data, are there any measures to address these? Information control: From the outset, information must be controlled, and the choice of which AI to use must be made. There are many AIs, mostly from the US and Europe, but the majority are outside the EU. Some tools are harmonized, depending on the user's choice and system. Judge's obligations: At what stage is there an obligation to disclose information when using AI? To whom are they accountable? There is no accountability, no sanctions, and no possibility of being sued for using AI. If problems arise, it is considered a lack of competence on the part of the judge. If a judge does not provide clear and thorough explanations, are they allowed to use AI? Why is AI applied to criminal cases first, instead of civil cases? Starting with AI in all fields, proving effective in large-scale disputes, those involving detailed damage calculations, and repetitive situations, AI is being prioritized for use in these civil cases. Accountability for AI needs to be emphasized in all areas, not just in the judicial system, because flawed AI can cause harm by making innocent parties appear guilty.
* Ms. Nguyen Thi Huyen Trang, lecturer at the Judicial Academy, posed a question to the French experts: The process of using AI by judges has high expectations in terms of AI training. When AI analyzes a judge's previous judgments and situations, what is the scope of such prohibition, and are there limitations on the AI's self-learning capabilities and the tools for controlling such prohibitions? * French expert, Ms. Ferreira Laisa, explained: When using AI in France, judges can be influenced by other EU decisions and rulings to avoid judges choosing their own approach based on how other judges have ruled. Input data for AI should not be carefully studied, for example, without prior research on legal doctrines and precedents, to allow for proper organization and analysis so that the AI ​​can identify and prioritize which legal provisions to apply. Judges and court clerks train the AI ​​based on the input data provided by the person responsible, not the judges themselves. Judges and clerks are the users. The AI ​​must accompany the judge throughout the entire litigation process, offering different legal options and bases, and the judge is responsible for the final choice. The AI ​​will guide the judge in applying the law; it only works with the judge, and it cannot be determined whether the AI ​​has been trained by the judge.
* French expert, Mr. Allain Matthieu, shared:
- AI training model at the French Court of Cassation: AI training at the Paris Court of Appeal, the Court of Cassation, and the Supreme Court. The AI ​​uses open-source code provided by a large private company in France, and there are other models provided by universities. The model is introduced into the Court of Cassation and adapted to the needs of the Court, ensuring two criteria: quantity and quality, so that the AI ​​becomes familiar with, for example, the structure and format of documents. The definition phase: experts structure the AI ​​themselves, specializing it in the field of justice. Three years ago, this was considered a large-scale language model. However, currently, it is only considered a small-scale model, oriented towards an open model but with specialization for AI according to individual preferences, and the N&M algorithm model only selects the appropriate AI model. Regarding the question of how the position of AI is reflected in the French Code of Professional Conduct?
- The Code of Conduct is not legally binding but is merely an internal administrative document of the Supreme Court, although it is regularly updated. From the beginning of 2025, there are intentions to incorporate AI into the Code, focusing on the use of specialized AI and open AI. It is hoped that regulations will be included in various relevant legal documents to raise awareness among those working with AI. Information entered into open AI could be leaked by some judges, requiring regulations to address such actions. Specific regulations will soon be codified under French law. Regarding the question of prohibiting automated judgments, how will the case of a judge using a judgment entirely drafted by AI be handled?
- Currently, there are no regulations assigning responsibility to AI. In all cases, responsibility rests with the judge, as if the judge hired someone else to execute the judgment. The process of holding a judge accountable is similar to when a judge hires someone else to create a judgment. While AI is widely used in drafting judgments, the judgment remains the outcome of justice. Therefore, if the result isn't the judge's own thinking, it violates citizens' rights to access a judgment rendered by a human.
* Mr. Pham Nhu Hung, former Deputy Director of the Judicial Academy, asked:
- Regarding Judge Cuong's question about checking the quality of input data, who controls the quality of input data? The experience of the Republic of France is to use specialized AI rather than open AI. The Vietnamese court's virtual assistant system collaborates with the military telecommunications group Viettel. In South Korea, 2400 IT professionals are a model for Vietnam to adopt a different approach to reduce the workload of Vietnamese judges.
- Is the AI ​​algorithm controlled by the French courts or by a private entity in France? The technological capabilities and quality of AI systems are developing. Currently, the French Republic has a team of court-affiliated technologists who can control input and output data and prevent hacking or data leaks. Recruiting a strong team of IT professionals with legal expertise is costly but proves more effective than lacking a specialized team. How is the law being adapted to work and litigate in a digital environment? - Vietnam often uses "judicial prediction." The advantages and disadvantages of open data in the French Republic allow for predictive judicial action, but this is not provided by judicial bodies; rather, it relies on public AI tools.
 
- Predictive Justice: For the past 10 years, there have been criticisms of software that analyzes the recidivism rate of citizens. It's necessary to distinguish between "predictive justice" in the general sense that people understand and data sources used to help citizens understand the progress of disputes. Justice can be predictive in nature, not just predictive as before.
- Using AI benefits not only judges but also citizens. In France, there is training for citizens in general on accessing digital transformation, especially in judicial activities. What benefits do citizens gain from the modernization of the justice system through the application of AI?
- To effectively communicate information to the public, citizens need to understand why a particular judgment was issued. In judicial proceedings, policies are geared towards administrative and judicial modernization. Since 2014, regulations concerning electronic signatures in general, not just within the judicial sector, have been in place. Electronic signatures are guaranteed to be secure, meet standards, and ensure the signature accurately represents the individual.
- The increasing number of digital platforms in the judicial system in recent years has led to the digitization of documents and procedural processes, both currently and in the future, towards complete digitization of trials and documents in digital data format. Individuals are held accountable for the publication of judgments and the use of digital signatures. A digitized judgment is recognized only when it bears the digital signature of the judge, clerk, and seal. Should we stick to the old logic or shift to the new logic of digitalization? But how can we completely eliminate paper versions? There are currently no specific legal regulations to address this.
Through presentations and information exchange by experts, attendees gained knowledge and learned about the skills of judges when applying AI, ensuring transparency and professional ethics when using AI in resolving civil cases in France and Vietnam. This contributed to improving the quality and effectiveness of teaching sessions on knowledge and skills related to the application of artificial intelligence in judicial activities for judicial positions at the Academy. The workshop successfully completed its program and content, meeting its objectives and requirements; ensuring compliance with the regulations and the Ministry of Justice's foreign affairs management regulations.
2. Evaluation of Advantages
The workshop's content received significant interest and participation from delegates in the context of the ongoing Fourth Industrial Revolution, where artificial intelligence (AI) has had a profound impact on all aspects of social life, including the judicial sector. Organizing the workshop online via Zoom facilitated connectivity between Vietnam and the Republic of France, allowing delegates from different provinces and cities to participate online. The workshop's organization was meticulous, ensuring adequate facilities and technical support. The above is a report on the results of the workshop on "Skills and Experiences of Judges in Applying Artificial Intelligence in Resolving Civil Cases," which the Academy respectfully submits to the Ministry of Justice leadership.