Abstract
In the legal field, robot judges and lawyers often cause people's concerns, especially the breakthrough development of generative artificial intelligence, which has triggered a new round of anxiety about machines replacing people. Legal artificial intelligence systems represented by ChatGPT4.0 and Deepseek have attracted the attention of the legal community. As legal technology, they provide a specific practical picture. Legal artificial intelligence first breaks through the cognitive limitations of lawyers and realizes rational decision-making and knowledge innovation. They focus on complex legal reasoning activities, reduce costs, improve efficiency, break the traditional growth model of the master-apprentice relationship, and are conducive to the balanced distribution of legal service resources and the improvement of international participation. Unavoidable problems arise with it. Compared with the legal services provided by traditional lawyers, legal artificial intelligence lacks empathy and value judgment ability, the reliability of the data models and analyses generated by algorithms they provide is questionable. The differences and imbalances in the input of the original data will make it difficult to guarantee the authenticity and security of the information. Finally, the responsible subjects and supervision of legal artificial intelligence are still unclear. Taking the legal services provided by lawyers as the clue, using comparative analysis, empirical analysis, and case analysis, the advantages and disadvantages of traditional legal profession and artificial intelligence lawyer skills are compared. In the future, human-machine symbiosis and human-machine interaction will be common phenomena. Making full use of the ever-evolving technology to improve the accuracy of lawyers in their practice areas is a trend of integrating high technology and professional knowledge.
Keywords
Artificial Intelligence, Legal Technology, Algorithm, Collaborative Governance, Human-computer Interaction, Pragmatism
1. Introduction
The concept of artificial intelligence (AI) first appeared in August 1955, mathematician John McCarthy, computer and cognitive scientist Marvin Minsky, IBM system designer, and information theory founder Claude Shannon discussed, “we will try to discover how to make machines use language, propose abstract propositions and concepts, solve certain problems currently left to humans, and achieve self-improvement”
, AI has been a cause that a group of like-minded scientists and engineers have been pursuing since the beginning, its purpose is to enable machines to learn and complete tasks that humans need to use intelligence to complete. It is an innovation of the traditional programming model. This cause has not been smooth sailing. It has experienced several winter, until recently it was considered a core technology of the Fourth Industrial Revolution, with the potential to revolutionize the human condition
.
With the birth of ChatGPT in November 2022 and Deepseek in January 2025, the powerful computing power and adaptive learning ability demonstrated by large-scale language models and pre-training technology enable it to simulate the creative thinking process of humans and generate more complex and realistic content. Legal artificial intelligence (LAI) relies on the underlying technical framework of artificial intelligence algorithms and uses automation technology to complete prosecution, trial and other work that is usually done by legal professionals with professional knowledge. It has now become the forefront of legal technology to provide practical legal service solutions
. It is oriented to big data, legal theory and legal practice, and uses algorithmic logic to match the increasingly complex and diverse legal practice needs on the basis of following the laws of legal operation, assisting the legal profession to make reasonable decisions and improve the quality and efficiency of legal work.
More and more traditional legal service platforms have been upgraded to artificial intelligence systems that can dynamically adapt to and respond to complex legal issues in real time, and realize harmonious human-computer interaction under the empowerment of technology, and have gradually formed legal artificial intelligence forms such as electronic dataization of information
, intelligent case handling assistance systems
, entity judgment prediction and supervision systems
, and electronic evidence standards
| [6] | Liu Xieci and Luo Shiting (2024). Legal artificial intelligence empowers lawyers' practice: prospects, obstacles and approaches. Journal of Nan Ning Normal University (Philosophy and Social Sciences Edition), vol. 45, no. 6. pp. 97-121. https://doi.org/10.16601/j.cnki.issn2096-7349.2024.06.008 |
[6]
. Because legal artificial intelligence has a strong ability to process natural language, it can actively transform massive pre-trained legal data into coherent dialogue texts that meet user needs, so it is favored by legal practitioners at home and abroad.
Although with law as the background, it not only attempts to provide differentiated legal services for lawyers, the public, enterprises, judges, prosecutors and other subjects, but also produces research and development results with certain practical application capabilities in different micro-business scenarios
, such as similar case recommendations, legal document generation and review, judgment result prediction, litigation risk analysis, and warning of different judgments
. However, risk would come to dominate legal practice and lead to a structural collapse of the legal profession.
Taking legal skills as the analysis clue, the comparative methodology applies to the differences and changes between traditional legal profession and lawyers in the AI’s era, and uses a pragmatic methodology to analyze the dialectical relationship between LAI and lawyer skills in the near future.
2. LAI’s Scenarios in Lawyers’ Practice
The legal services provided by lawyers are essentially the mapping of client needs with relevant legal knowledge, combined with practical experience, to help clients resolve complex social relationships. LAI, with its powerful embedding, understanding and generation capabilities, is expected to accurately match existing legal provisions with case facts. LAI empowering lawyers is expected to break the traditional growth model and accelerate the transformation and upgrading of internal bureaucratic structure and external market structure. LAI has the possibility of evenly distributing legal service resources and increasing international participation.
2.1. Breaking Through Cognitive Limitations
Knowledge linkage ability and unlimited memory ability are the LAI’s core elements, successfully empower lawyers to practice. In the diversified society, clients services come from all walks of life and their needs are even more varied, this requires not only lawyers to master solid professional knowledge, excellent legal document writing, legal case retrieval, case analysis and other basic abilities, but also to be good at applying various professional knowledge and abilities to case handling
| [15] | Xu Jun and Ding Jianing (2025). Challenges of judicial trust and the construction of trustworthy standards in the intelligent era. Journal of China Three Gorges University (Humanities and Social Sciences), vol. 47, no. 2, pp. 78-83. https://doi.org/10.13393/j.cnki.1672-6219.2025.02.013 |
[15]
. The powerful computing power and memory ability in LAI are highly compatible with the needs of legal services. Lawyers have high legal and technical literacy due to systematic legal education and technical training, and can more efficiently realize the integration and embedding of legal business, case and technology
| [14] | Xie Xinshui (2025). Intelligent transition, open-source innovation and sovereign AI: Deep Seek phenomenon promotes the construction of open-source innovation ecosystem for artificial intelligence. E-Government, vol. 03, pp. 40-48. https://doi.org/10.16582/j.cnki.dzzw.2025.03.002 |
[14]
.
The theory of bounded rationality holds that decision makers are easily dominated and disturbed by incomplete information in the decision-making process and can only make limited decisions within their capabilities
| [15] | Xu Jun and Ding Jianing (2025). Challenges of judicial trust and the construction of trustworthy standards in the intelligent era. Journal of China Three Gorges University (Humanities and Social Sciences), vol. 47, no. 2, pp. 78-83. https://doi.org/10.13393/j.cnki.1672-6219.2025.02.013 |
[15]
. When lawyers start to participate in legal work, they inevitably need to meet with clients, answer questions and provide legal solutions. Their decision-making ability is often one of the criteria for clients to evaluate. LAI has the ability to extract and mine meaningful knowledge and form models from massive legal knowledge in an efficient and refined manner according to instructions, and then summarize and output information about case predictions, litigation processes, legal strategies, etc., and analyze the winning and losing trends and defense models in previous litigation, so as to break through the limitations of lawyers' knowledge linkage and alleviate the limitations of legal decision-making.
Another aspect is, LAI helps lawyers overcome the limitations of memory. Common law countries in the United States and the United States applied LAI before codified law countries because the technology compiles, classifies and queries the vast case files to meet the normal operation of the national judicial system. However, in codified law countries, lawyers, especially junior lawyers, face a complex legal system and an increasing number of laws, regulations and judicial interpretations. They urgently need the powerful memory and retrieval capabilities to help them get out of the dilemma of manpower and time costs being disproportionate to the comprehensiveness of memory and retrieval. The powerful functions can make up for some limitations of human intelligence, greatly liberate the mental labor of lawyers, and enable them to concentrate on more complex legal reasoning activities.
2.2. Breaking the Traditional Growth Model
LAI breaks the traditional “professor-apprentice”
growth model of lawyers and helps upgrade the internal bureaucratic structure and external market structure of the industry. The legal industry mostly adopts an elite inheritance system. From the analysis of the internal hierarchical structure, a professor with rich practical experience plays a vital role in the rapid growth of lawyers. This means that legal knowledge acquisition, legal analysis and reasoning training are the products of the power difference between professor and apprentice, which has caused a strong personal dependence relationship, and caused the internal hierarchy to solidify into a case-intensive structure centered on high-level lawyers
| [14] | Xie Xinshui (2025). Intelligent transition, open-source innovation and sovereign AI: Deep Seek phenomenon promotes the construction of open-source innovation ecosystem for artificial intelligence. E-Government, vol. 03, pp. 40-48. https://doi.org/10.16582/j.cnki.dzzw.2025.03.002 |
[14]
.
With the in-depth LAI’s application, the above problems have been alleviated to a certain extent. On the one hand, the technology weakens the time constraints of legal knowledge accumulation and legal reasoning ability acquisition, and eases the dependence of junior lawyers on senior lawyers
. The generation and accumulation of legal knowledge requires a long period of legal hermeneutics research and training to achieve the transformation of legal information into legal knowledge. Legal analysis and reasoning need to follow the rules of experience, which is essentially a process of following precedents. In China's judicial practice, the principle of “similar judgments for similar cases” can be implemented not because similar cases have absolute factual similarities
| [15] | Xu Jun and Ding Jianing (2025). Challenges of judicial trust and the construction of trustworthy standards in the intelligent era. Journal of China Three Gorges University (Humanities and Social Sciences), vol. 47, no. 2, pp. 78-83. https://doi.org/10.13393/j.cnki.1672-6219.2025.02.013 |
[15]
, but because of their reference in reasoning logic.
However, it is not easy for lawyers to master the methods of legal analysis and reasoning. They must undergo a lot of logic training to master and apply them proficiently. The deductive reasoning logic of legal artificial intelligence is based on legal rules, cases and similar facts, and evolves from legal propositions to legal conclusions with the help of non-single algorithm operation paths
, which is similar to the thinking logic deduction form of lawyers’ legal services. LAI can retrieve cases with similar reasoning processes and provide lawyers with less experience with a more comprehensive and sophisticated legal reasoning model. On the other hand, LAI is a limited means to change the traditional bureaucratic structure of law firms. Generally speaking, senior lawyers are in a central position within law firms because of their rich professional experience and abundant business resources, but junior lawyers have more potential in learning and using, because of their wide inclusiveness and strong acceptability. From the analysis of the external market structure, LAI drives innovation in external market services. Due to the saturation of the traditional legal service market and the rise of legal technology companies, to a certain extent, lawyers have used the wisdom of LAI to develop pre-risk prevention and control business.
2.3. Balancing Service Resource Distribution to Enhance International Participation
LAI solves the uneven distribution of lawyer service resources and provides an opportunity for Chinese lawyers to deeply participate in the international legal service market. Due to economic development, geographical location, political culture and other reasons, Chinese lawyer service resources are mainly distributed in the eastern coastal areas, while the central and western regions are relatively scarce. The training of foreign-related lawyers, young lawyers, county lawyers and other talents is tilted towards the underdeveloped western regions, aiming to solve the problem of unbalanced development in the east, middle and west
. The design and development of LAI strives to deeply integrate AI with judicial big data to promote the equalization and accessibility of legal services and make up for the gap in the enjoyment of legal service resources between urban and rural areas, regions and different groups of people
. This value pursuit is highly consistent with the purpose of my country's lawyer service resource allocation policy, or it can temporarily fill the gap in legal talents. On the one hand, convenient and efficient LAI can provide individuals or organizations with legal consultation, document review, case prediction and other services at low cost
| [11] | Wei Yuming, Jia Kai and Zeng Runxi (2025). Innovation and development of artificial intelligence and governance reform under the breakthrough effect of Deep Seek. E-Government, vol. 17, no. 3, pp. 2-39. https://doi.org/10.16582/j.cnki.dzzw.2025.03.001 |
[11]
; on the other hand, the massive data storage capacity and deep learning ability of legal artificial intelligence can quickly learn and digest the regional laws and social norms of the underdeveloped areas in the central and western regions to adapt to special circumstances, which is conducive to the output of more accurate and reasonable legal solutions
| [8] | Ning, Li-ang (2025). Three questions on criminal liability involving artificial intelligence: the object of reference, infringement of legal interests and the subject of responsibility. Journal of Zhe Jiang University, vol. 02, pp. 64-73. https://doi.org/10.16235/j.cnki.33-1005/c.2025.02.024 |
[8]
. In addition, the deep application of technology helps junior lawyers to correct and improve their knowledge system and practical experience in real time during the case trial process, so that lawyers in different regions and levels can make up for the mistakes that may occur due to differences in knowledge structure and the amount of case judgment experience.
As Chinese foreign-related business market is constantly expanding, the demand for foreign-related lawyers is also further expanding. Foreign-related lawyers not only need to be familiar with domestic laws and regulations, international commercial treaties, international practices and foreign civil and commercial legal knowledge, but also need to have foreign language skills that are superior to ordinary people and have a comprehensive and reasonable legal knowledge structure. Legal language is highly complex, which is reflected in many specialized terms. It is highly formulaic and academic and has a semantic gap with everyday language. Seemingly precise legal terms may also have multiple meanings, such as the same term may have different, context-specific meanings in legal language. LAI’s users can conveniently obtain professional knowledge in foreign law, finance, medical and other fields, as well as linguistic knowledge that breaks through natural language processing technology, to achieve smooth interaction and systematic transformation between natural language and legal language, thereby assisting lawyers to participate deeply in the formulation of international trade rules that meet the actual cooperation needs in the international arena.
3. LAI’s Real Obstacles in Lawyers’ Practice
Although LAI brings many technical benefits to lawyers with the support of data, algorithms, and computing power, value judgment capabilities, it lacks empathy and the reliability of data models is questionable, the reliability and security of collected and produced information is difficult to guarantee, and it is difficult to identify responsible and regulatory entities.
3.1. Lack of Empathy
Machines cannot accept or express emotions, nor can they have actual emotional connections with people. Lawyer services not only need to provide basic legal services, but also need to integrate guidance, consultation, emotional support and other services for clients in the business. Lawyers need to patiently listen and guide clients to describe the specific case, analyze the relevant legal relationships and transform the case facts into legal facts, and finally determine the specific cause of the case and file a lawsuit.
The fact that must be faced is that most parties, due to lack of legal knowledge, mostly express their opinions from a subjective perspective. Lawyers also need to capture the emotions and inner sincerity of the parties through subtle observations, guide the parties to express their true demands and formulate targeted litigation strategies. Most of these cases are in marriage, inheritance and other aspects
. In civil cases involving more human emotional issues, the function of LAI is not powerful. Therefore, although LAI has a strong ability of deep autonomous learning, it cannot acquire empathy and value judgment unique to humans soon. AI represents the cold side of the law, while artificial lawyer consultation has the warmth of humanity
.
3.2. Uncertainty About Data Reliability
LAI is generally used by self-deep learning of existing legal data and materials, and by mobilizing legal knowledge to creatively generate legal conclusions according to the instructions issued by lawyers. However, the stability and reliability of the underlying model of existing data operations still need to be strengthened, which is mainly affected by subjective and objective reasons. From a subjective perspective, differences in the subjective values of developers affect the reliability of data models
. The formation of basic models requires three stages, namely the collection and screening of model data, the training and reward model stage, and the manual inspection of model reliability
. These stages are inseparable from human intervention. Due to differences in the values of developers, bias, discrimination, errors, and incompleteness are prone to exist in the data, resulting in deviations or even misleading output content of specific legal data models, which may cause irreparable losses to users. From an objective perspective, the richness of legal knowledge affects the stability of data models. The diversification of legal sources and the regionalization of legal content make legal knowledge rich; LAI cannot fully meet the requirements of the richness of legal knowledge, this is mainly because the large-scale legal data crawled by LAI cannot be equated with the richness of legal knowledge.
From the perspective of data production, the amount of data contributed by young people and users in developed countries is significantly higher than that of the elderly and developing countries
. At the same time, people who hold mainstream values and hegemonic views can easily produce a large amount of data, while other people with insufficient representation are just the opposite
. This makes it impossible for LAI to cover a large amount of implicit knowledge and tacit knowledge in the process of knowledge absorption and reproduction, and it is also unable to identify and absorb local legal knowledge.
3.3. Insufficiency of Data Security
LAI is limited by its own technical functions and cannot guarantee the reliability of its output results. The data collection of LAI is limited and has a large amount. From a technical principal point of view, legal AI uses a large amount of labeled high-quality legal data as simulation training corpuses, but the reality is that the total amount of data in the legal field is very limited, and most of these legal data are presented in an unlabeled and unstructured form
, which is difficult for computers to directly process automatically. As a result, the legal analysis based on pre-training data cannot cover the latest legal data and information
| [14] | Xie Xinshui (2025). Intelligent transition, open-source innovation and sovereign AI: Deep Seek phenomenon promotes the construction of open-source innovation ecosystem for artificial intelligence. E-Government, vol. 03, pp. 40-48. https://doi.org/10.16582/j.cnki.dzzw.2025.03.002 |
[14]
, has a lag, and may produce misleading legal opinions, and cannot meet the practice needs of lawyer users who try to obtain the latest legal knowledge through LAI assistants.
The generation function of LAI is based on data integration, but it will not automatically screen data and judge it true or false, resulting in the reliability of the generated results being questioned. For example, during the data acquisition stage, OpenAI collected hundreds of billions of information data resources without the consent of the data holder, including a large amount of personal sensitive information. Although it will take desensitization measures such as anonymization to prevent information leakage for non-public data or private information, it could integrate fragmented information into complete personal information, and even outline personal portraits, which poses a risk of personal privacy leakage
.
In addition, in the process of interacting with lawyer users, to achieve accurate services, legal artificial intelligence requires users to actively provide detailed information including preferences, personal conditions, etc., and it can grasp more user personal information with its powerful storage capacity. This information may be shared or sold among industries when the technology is redeveloped and upgraded, resulting in personal privacy infringement.
4. LAI’s Regulation and Development in Lawyers’ Practice
Faced with the future trend of legal artificial intelligence technology being deeply embedded in the legal profession, improvements should be made in the following four aspects.
4.1. To Shape the Interaction of Human and AI
The harmonious symbiotic relationship between lawyers and technology should be taken seriously. The values and emotional judgment ability of lawyers should be deeply combined with the high efficiency of technology to shape the interactive mode between industry people and technology, help liberate brain power and optimize the practice mode of lawyers.
Through the above analysis, it is not difficult to find that lawyers and LAI have their own strengths in solving legal problems. Lawyers can make value judgments and value reasoning, so that the solution of legal problems can adapt to the rapid changes of society and have flexibility. LAI has strong memory and retrieval capabilities, can quickly and comprehensively analyze and solve legal problems, and has stability. The realization of the envisioned plan is inseparable from the efforts of lawyers and technical developers.
Starting with the lawyers themselves, improving the digital technology literacy of lawyers has become the primary task, to maximize LAI’s advantages, it is necessary to cultivate lawyers' new technical literacy, especially the technical literacy of structured legal prompts, based on maintaining the subject status of legal persons. Only when lawyers have mastered the ability to accurately describe and solve legal problems and comprehensively use prompting technologies such as legal provisions, legal cases, and legal explanations can they output high-quality answers.
In addition, from a technical level, injecting legal thinking into LAI can empower the development of the legal industry. Although consciousness is a product unique to humans, due to the high degree of simulation and deep learning capabilities, developers can classify information during the data collection stage and deliver content with subjective emotions and value judgments to the machine. The machine generates content output after forming a certain legal thinking mode through autonomous learning. When obtaining legal information with subjective thoughts such as legal judgment documents and lawyers' defense statements, the algorithm technology will store this part of the information separately and analyze and generate a case-like subject thinking model, which is more in line with the lawyer's case-handling thinking and reduces the lawyer's reprocessing cost of generated content.
4.2. To Optimize the Text Generation Mechanism and Standardize the Algorithm Model
LAI’s core is the basic algorithm model and text generation mechanism, but the problems of algorithm bias and algorithm discrimination still exist. The reliability and security of basic data need to be strengthened. Correcting the deviations of the basic algorithm model and text generation mechanism built with massive legal data is of great significance for lawyers’ practice. Different ports in the mechanism may lead to different degrees of data security risks, different measures should be taken for different ports
.
As far as the input end is concerned, first of all, strictly control the passive outbound flow of data. The network is convenient and fast, and it is difficult to predict the flow of data. Active defense measures should be taken, such as establishing an overseas network attack monitoring and dispatching platform to implement unified management and monitoring of important data in the lawyer industry. Secondly, to strengthen the market access rules for legal artificial intelligence. We should be cautious about LAI products that request entry. We must clarify the security assessment rules and corpus data acquisition rules. We should also require R&D institutions to adopt data cleaning and de-identification technical means to ensure the accuracy and security of data information and ensure that the foundation of the data model is stable.
On the computing side, R&D personnel should work with senior lawyers, judges and prosecutors to increase the transparency, explainability and accountability of algorithm operations
. Specifically, based on the characteristics of algorithms that are difficult to explain and non-intuitive, the principle of algorithm transparency reveals the value judgments behind the output text of LAI, so that the algorithm reveals its true face and is accountable. In addition, it is also necessary to do a good job of post-event response and punishment, especially to promote the establishment of reporting and rumor-refuting mechanisms on technology platforms
, and to take restrictive measures such as stopping transmission for those who illegally spread false and harmful information. Senior lawyers can give some feasible suggestions on the design of post-event accountability plans, which is convenient for specific applications later.
As for the output end, there may be a wide range of malicious content generation risks, and it is urgent to optimize the governance of the content generation system. Further research on algorithms should be promoted to ensure that the data generated by them meets ethical and legal standards. Lawyer users can feedback maliciously generated content to the legal technology service platform. The platform improves its ability to monitor, filter and block malicious content by training and adjusting models, prevents the emergence of malicious content, and establishes a blacklist mechanism to record and manage malicious content and its disseminators. The industry needs to formulate standardized technical standards and implement standardized management from the research, development, and application of LAI, so as to improve technical security and reliability, promote the generated content to comply with public order and good customs and legal provisions, and meet the needs of lawyer users.
4.3. To Control Data Reliability and Security
High-quality legal services depend on effective protection of data reliability and security. Whether it is a pre-case meeting or an in-case defense, lawyers can rely on LAI system for auxiliary work, which involves a large amount of legal data and personal data. This requires that in the service process, we should not only pay attention to the reliability of collecting, screening, summarizing and generating data, but also effectively protect data security and improve the security of data privacy. From the perspective of ensuring data reliability, the goal is to improve the transparency of Generative AI (GAI) services and improve the accuracy and reliability of generated content. However, it is undeniable that current technology is still unable to comprehensively screen and verify the authenticity of massive data in LAI’s database. Due to the autonomous learning ability and text generation model, even if all the database information is true, it is impossible to eliminate the LAI’s possibility to generate erroneous information. Therefore, it is necessary to solve LAI to generate erroneous information. First, LAI’s source and quality should be further guaranteed to ensure that the collection and processing of data strictly follow legal norms and technical standards. Strict supervision and management should also be implemented throughout the process of data processing and analysis to ensure the reliability of data. Secondly, to clarify the responsibility-bearing measures is an effective means to solve the problem of false data generated by the system.
4.4. To Improve Collaborative Governance and Regulatory Rules
Governance requires the participation of multiple parties, clarifying the responsibilities and focus of each party in order to fill the regulatory gap. The generation process of legal artificial intelligence technology mainly involves three groups: the state, the platform and the user. The platform includes information collection platform, research and development platform and service provision platform. It is unrealistic to rely on the power of one party for governance and supervision. Therefore, all parties need to reach a consensus on the governance and supervision goals and systems, perform their respective duties, and ensure that the technology can play its maximum role in lawyers' practice safely, efficiently and conveniently.
In China, normative documents represented by the “Interim Measures for the Administration of Generative Artificial Intelligence Services”, “Regulations on the Ecological Governance of Network Information Content”, “Guiding Opinions on Strengthening the Comprehensive Governance of Internet Information Service Algorithms”, “Regulations on the Administration of Internet Information Service Algorithm Recommendations”, and “Regulations on the Administration of Deep Synthesis of Internet Information Services” require relevant departments to ensure technological development, promptly handle user complaints or reports on services, and conduct inspections on services, order rectification and impose penalties on illegal services.
In the full-scale and full-process supervision of platforms and users, we must also maintain a high degree of sensitivity to the development of legal artificial intelligence technology and have a high ability to predict its possible infringements and harms. This undoubtedly puts higher demands on the supervision capabilities of relevant national departments, which will naturally increase administrative costs and may even hinder the progress of technology due to excessive intervention. Therefore, we should focus on developing legal artificial intelligence and improving the citizen rights relief mechanism.
LAI’s service platforms should actively improve industry technical specifications and standards. Some scholars believe that soft laws such as industry self-discipline conventions, ethical norms, and standard guidelines are more flexible and adaptable, can serve a variety of governance goals, and have become the most common form of artificial intelligence governance. However, with the continuous development of legal artificial intelligence, various technical specifications and standards, as well as industry norms such as technical ethics of industry associations, are no longer in line with the development stage. Therefore, each platform should focus on standardizing the technology and industry standards of each operating stage of legal artificial intelligence, including but not limited to the norms in the work process of each platform and the norms for the connection between platforms. For example, the standards followed by the information collection platform for judging, screening and filtering malicious content should be updated with the changes of the times.
Lawyers should improve their own technical control capabilities and consolidate the foundation of legal professional knowledge, cultivate the ability to identify generated content, and actively feedback malicious and erroneous knowledge content to the platform, help the platform improve the information feedback and technology optimization system, and achieve a qualitative leap
| [6] | Liu Xieci and Luo Shiting (2024). Legal artificial intelligence empowers lawyers' practice: prospects, obstacles and approaches. Journal of Nan Ning Normal University (Philosophy and Social Sciences Edition), vol. 45, no. 6. pp. 97-121. https://doi.org/10.16601/j.cnki.issn2096-7349.2024.06.008 |
[6]
in technical content through dynamic adjustment. According to the user agreement, the correct use of technology to empower lawyers to practice, and shall not use technology to induce others to commit crimes and infringe on the rights and interests of others. Finally, lawyers should abide by the leading role in legal knowledge verification and legal knowledge innovation. Since law itself is an intellectual activity with critical thinking as its core, lawyers are essentially required to maintain a leading role in legal practice. Therefore, in the face of diversified practice scenarios, lawyers should look at the generated content of legal artificial intelligence dialectically, use critical thinking to think independently and make self-calibrated legal knowledge judgments, and draw reasonable legal conclusions.
5. Conclusion
LAI provides powerful technical means for legal practice. It cuts into cognitive patterns, behavioral costs, growth paths, internal and external structures, resource allocation and international participation at the same time, realizing comprehensive and efficient technical capabilities for the legal field, and further transforming the traditional legal service model based on experience into a digital legal service model based on technology.
In the future, we need to avoid the digital divide and achieve further iteration of legal artificial intelligence by strengthening the supply of high-quality legal data and building a dynamic adjustment mechanism based on judicial big data and a reinforcement learning mechanism based on knowledge feedback from lawyers. The lawyers' practice group and the technology development group need to go hand in hand, organically integrate the evolution law, technical logic and legal spirit and legal knowledge, optimize the text generation mechanism and algorithm model, improve the collaborative governance and supervision path, and promote the new process of LAI empowering lawyers' practice on the basis of maintaining the independent personality and emotional thoughts of lawyers. When the legal profession group works together, continuously and dynamically adjusts artificial intelligence to understand and follow the practice law of lawyers, conforms to the practice characteristics, locates and plays the role of assistants, and maximizes the technical advantages of legal artificial intelligence, can we meet the scene needs of legal practice under the premise of following the practice law of lawyers, assist lawyers in making accurate legal decisions, and comprehensively improve the quality of lawyers.
Abbreviations
AI | Artificial Intelligence |
LAI | Legal Artificial Intelligence |
GAI | Generative Artificial Intelligence |
Author Contributions
Yang Yang is the sole author. The author read and approved the final manuscript.
Funding
This paper is supported by the University of Alicante, Spain, as the part of the Programa Redes-I3CE en Investigación y Docencia Universitaria-6298-Universidad de Alicante.
Conflicts of Interest
The author declares no conflicts of interest.
References
| [1] |
Fu Jianyu (2025). ChatGPT-like large model empowers proportional control of digital government. Science and Technology and Law, vol. 22, no. 2, pp. 63-74.
https://doi.org/10.19685/j.cnki.cn11-2922/n.2025.02.007
|
| [2] |
He Donghang (2025). Research on artificial intelligence and Chinese modernization: An investigation based on the current situation, challenges and countermeasures. Journal of Social Science, vol. 03, pp. 1-14.
https://kns-cnki-net.webvpn.synu.edu.cn/kcms/detail/21.1012C.20250305.0926.026.html
|
| [3] |
Li Xiaomeng (2023). Defense lawyers’ data rights in the era of big data: as evidence of emerging rights. Journal of Shanghai University of Political Science and Law, vol. 21, no. 4, p. 117-125.
https://doi.org/10.19916/j.cnki.cn31-2011/d.2023.04.006
|
| [4] |
Li Zhi and Luo Xing (2025). Human resource reconstruction under the intelligent revolution: Research on the challenges, impacts and governance of Deep Seek and Manus-like generative artificial intelligence on the human resource market. Journal of Chongqing University (Social Science Edition), vol. 21, no. 3. pp. 1-13.
https://kns-cnki-net.webvpn.synu.edu.cn/kcms/detail/50.1023.C.20250311.1349.007.html
|
| [5] |
Liu Hao (2025). The logic and approach of legal protection of credit data in generative artificial intelligence. Science and Technology and Law, vol. 22, no. 2, pp. 103-113.
https://doi.org/10.19685/j.cnki.cn11-2922/n.2025.02.011
|
| [6] |
Liu Xieci and Luo Shiting (2024). Legal artificial intelligence empowers lawyers' practice: prospects, obstacles and approaches. Journal of Nan Ning Normal University (Philosophy and Social Sciences Edition), vol. 45, no. 6. pp. 97-121.
https://doi.org/10.16601/j.cnki.issn2096-7349.2024.06.008
|
| [7] |
Lu Feng (2025). Research on the deep integration, development and application of artificial intelligence technology and legal methods. Western Journal, vol. 04, pp. 49-52.
https://doi.org/10.16721/j.cnki.cn61-1487/c.2025.04.022
|
| [8] |
Ning, Li-ang (2025). Three questions on criminal liability involving artificial intelligence: the object of reference, infringement of legal interests and the subject of responsibility. Journal of Zhe Jiang University, vol. 02, pp. 64-73.
https://doi.org/10.16235/j.cnki.33-1005/c.2025.02.024
|
| [9] |
Sun Guorui and Liu Weibin (2025). Technical governance and legal governance of artificial intelligence. Science and Technology and Law (Chinese and English), vol. 13, no. 2, pp. 1-10.
https://doi.org/10.19685/j.cnki.cn11-2922/n.2025.02.001
|
| [10] |
Tang Yang, Dong Xiaoying (2025). Global artificial intelligence regulatory trends and my country's response strategies. Communications World, vol. 34, no. 01, pp. 6-7.
https://doi.org/10.13571/j.cnki.cww.2025.04.003
|
| [11] |
Wei Yuming, Jia Kai and Zeng Runxi (2025). Innovation and development of artificial intelligence and governance reform under the breakthrough effect of Deep Seek. E-Government, vol. 17, no. 3, pp. 2-39.
https://doi.org/10.16582/j.cnki.dzzw.2025.03.001
|
| [12] |
Wu Guobang (2025). Legal construction of justice and human-machine relationship. Yunnan Social Sciences, vol. 23, no. 2, pp. 84-93.
https://kns-cnki-net.webvpn.synu.edu.cn/kcms/detail/53.1001.C.20250311.0824.012.html
|
| [13] |
Xie Yaowen (2025). Matching administrative supervision of the generative artificial intelligence value chain with tort liability. Political and Legal Forum, vol. 19, no. 01, pp. 1-11.
https://kns-cnki-net.webvpn.synu.edu.cn/kcms/detail/11.5608.D.20250227.1110.002.html
|
| [14] |
Xie Xinshui (2025). Intelligent transition, open-source innovation and sovereign AI: Deep Seek phenomenon promotes the construction of open-source innovation ecosystem for artificial intelligence. E-Government, vol. 03, pp. 40-48.
https://doi.org/10.16582/j.cnki.dzzw.2025.03.002
|
| [15] |
Xu Jun and Ding Jianing (2025). Challenges of judicial trust and the construction of trustworthy standards in the intelligent era. Journal of China Three Gorges University (Humanities and Social Sciences), vol. 47, no. 2, pp. 78-83.
https://doi.org/10.13393/j.cnki.1672-6219.2025.02.013
|
| [16] |
Zhang Wenxian (2025). On the deep integration of legal education and artificial intelligence. Digital Rule of Law, vol. 11, no. 3, pp. 1-8.
https://kns-cnki-net.webvpn.synu.edu.cn/kcms/detail/10.1879.d.20250306.1344.002.html
|
Cite This Article
-
APA Style
Yang, Y. (2025). Artificial Intelligence and the Future of Law -- Taking Lawyers’ Skills as an Observation. International Journal of Education, Culture and Society, 10(2), 93-100. https://doi.org/10.11648/j.ijecs.20251002.14
Copy
|
Download
ACS Style
Yang, Y. Artificial Intelligence and the Future of Law -- Taking Lawyers’ Skills as an Observation. Int. J. Educ. Cult. Soc. 2025, 10(2), 93-100. doi: 10.11648/j.ijecs.20251002.14
Copy
|
Download
AMA Style
Yang Y. Artificial Intelligence and the Future of Law -- Taking Lawyers’ Skills as an Observation. Int J Educ Cult Soc. 2025;10(2):93-100. doi: 10.11648/j.ijecs.20251002.14
Copy
|
Download
-
@article{10.11648/j.ijecs.20251002.14,
author = {Yang Yang},
title = {Artificial Intelligence and the Future of Law -- Taking Lawyers’ Skills as an Observation
},
journal = {International Journal of Education, Culture and Society},
volume = {10},
number = {2},
pages = {93-100},
doi = {10.11648/j.ijecs.20251002.14},
url = {https://doi.org/10.11648/j.ijecs.20251002.14},
eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ijecs.20251002.14},
abstract = {In the legal field, robot judges and lawyers often cause people's concerns, especially the breakthrough development of generative artificial intelligence, which has triggered a new round of anxiety about machines replacing people. Legal artificial intelligence systems represented by ChatGPT4.0 and Deepseek have attracted the attention of the legal community. As legal technology, they provide a specific practical picture. Legal artificial intelligence first breaks through the cognitive limitations of lawyers and realizes rational decision-making and knowledge innovation. They focus on complex legal reasoning activities, reduce costs, improve efficiency, break the traditional growth model of the master-apprentice relationship, and are conducive to the balanced distribution of legal service resources and the improvement of international participation. Unavoidable problems arise with it. Compared with the legal services provided by traditional lawyers, legal artificial intelligence lacks empathy and value judgment ability, the reliability of the data models and analyses generated by algorithms they provide is questionable. The differences and imbalances in the input of the original data will make it difficult to guarantee the authenticity and security of the information. Finally, the responsible subjects and supervision of legal artificial intelligence are still unclear. Taking the legal services provided by lawyers as the clue, using comparative analysis, empirical analysis, and case analysis, the advantages and disadvantages of traditional legal profession and artificial intelligence lawyer skills are compared. In the future, human-machine symbiosis and human-machine interaction will be common phenomena. Making full use of the ever-evolving technology to improve the accuracy of lawyers in their practice areas is a trend of integrating high technology and professional knowledge.
},
year = {2025}
}
Copy
|
Download
-
TY - JOUR
T1 - Artificial Intelligence and the Future of Law -- Taking Lawyers’ Skills as an Observation
AU - Yang Yang
Y1 - 2025/04/10
PY - 2025
N1 - https://doi.org/10.11648/j.ijecs.20251002.14
DO - 10.11648/j.ijecs.20251002.14
T2 - International Journal of Education, Culture and Society
JF - International Journal of Education, Culture and Society
JO - International Journal of Education, Culture and Society
SP - 93
EP - 100
PB - Science Publishing Group
SN - 2575-3363
UR - https://doi.org/10.11648/j.ijecs.20251002.14
AB - In the legal field, robot judges and lawyers often cause people's concerns, especially the breakthrough development of generative artificial intelligence, which has triggered a new round of anxiety about machines replacing people. Legal artificial intelligence systems represented by ChatGPT4.0 and Deepseek have attracted the attention of the legal community. As legal technology, they provide a specific practical picture. Legal artificial intelligence first breaks through the cognitive limitations of lawyers and realizes rational decision-making and knowledge innovation. They focus on complex legal reasoning activities, reduce costs, improve efficiency, break the traditional growth model of the master-apprentice relationship, and are conducive to the balanced distribution of legal service resources and the improvement of international participation. Unavoidable problems arise with it. Compared with the legal services provided by traditional lawyers, legal artificial intelligence lacks empathy and value judgment ability, the reliability of the data models and analyses generated by algorithms they provide is questionable. The differences and imbalances in the input of the original data will make it difficult to guarantee the authenticity and security of the information. Finally, the responsible subjects and supervision of legal artificial intelligence are still unclear. Taking the legal services provided by lawyers as the clue, using comparative analysis, empirical analysis, and case analysis, the advantages and disadvantages of traditional legal profession and artificial intelligence lawyer skills are compared. In the future, human-machine symbiosis and human-machine interaction will be common phenomena. Making full use of the ever-evolving technology to improve the accuracy of lawyers in their practice areas is a trend of integrating high technology and professional knowledge.
VL - 10
IS - 2
ER -
Copy
|
Download
Share This Article
Twitter
Linked In
Facebook
