AIGC-Enhanced Teaching of International Business Law Course in Business English Majors: Innovation and Practice ()
1. Introduction
Generative artificial intelligence has developed quickly in recent years. Large language models, which are often shortened as LLMs, excel among the most visible examples of this progress. They are changing industry and society in basic ways. Education has also felt this change. In March 2024, China’s Ministry of Education launched Leading Education through AI Development, an initiative also called LEAD. This initiative supports the creation of large models built specifically for education, one of which is Generative Educational Syntax Twin, known as GEST. It also put in place a clear policy framework for AI-supported modernization of education (Ministry of Education of China, 2024).
In the field of higher education, more and more scholars are studying how to use AIGC technologies in teaching. Internationally, Khan Academy built Khanmigo on the basis of ChatGPT, and Duolingo launched Duolingo Max, which runs on GPT-4 to help users learn languages. These two products show how much AIGC can help with language teaching (Xu, 2024). Domestically, Professor Xu Jiajin and his team at Beijing Foreign Studies University wrote a monograph talking about LLMs used in foreign language teaching and research. The book gave usable theoretical support for related work (Xu, 2024). Right now, not many people have studied how to combine AIGC with core Business English courses. Few studies even touch on courses that need both legal professional knowledge and strict language use; International Business Law is a typical example of this kind of course. No relevant discussion can be found in existing academic writing.
International Business Law fills a space in the Business English curriculum. It asks students to master both professional legal knowledge and advanced English skills at the same time. Students usually face several different kinds of difficulties when they study this course (Zhang & Ma, 2023). First, the range of legal knowledge involved is very wide. It covers contract law, the law of international sale of goods, intellectual property law and international investment law. Besides, another significant barrier is the legal English, whose specialized terms, complex sentence structures, and formal writing style are of great difference from general English. Most learners struggle to understand this content and produce related written or oral work. Third, the textbook of International Business Law rarely includes enough authentic English teaching materials and real case studies. This poses another challenge to the quality of teaching directly. Fourth, students rarely get chances to apply what they learn in practice. Activities like moot courts and client simulations that let students gain hands-on experience are not easy to access, which stops students from turning the theories they learn into practical professional skills.
The emergence of AIGC technologies provides a new way to address these challenges. When educators build intelligent instructional support systems, they can create high-quality English case materials, deliver personalized language coaching, build virtual practice environments, and develop multi-dimensional assessment frameworks. These changes raise the quality of IBL instruction and help students build stronger professional competencies.
This study works to connect AIGC technological capability with educational application, under the context of International Business Law teaching in the business English majors. It gives empirical evidence and practical guidance for curriculum innovation.
2. Literature Review and Theoretical Framework
2.1. AIGC Technologies and Their Educational Applications
AIGC (Artificial Intelligence-Generated Content) covers a group of technologies that use generative AI to create text automatically, and images, audio, video, as well as other multi-modal content. After ChatGPT was released to the public in November 2022, large language models like GPT-4 have shown unprecedented ability to understand and generate natural language, and opened up new space for use in education.
International organizations have seen more and more language educators around the world use AI tools in their daily work. The British Council (2024) reported that large-scale surveys of English teachers worldwide keep pointing out that AI can automate routine work, tailor learning to individual needs, and give learners instant feedback on their written or spoken work. In the specific area of language teaching, AIGC applications can be divided into four general functions. a) Intelligent tutoring systems can answer questions at any time and offer support for language practice. b) Content generation tools help teachers quickly put together a range of different instructional materials. c) Personalized learning partners can simulate real conversational contexts for learners to practice their speaking and writing. d) Assessment tools use natural language processing to give timely diagnostic feedback on how learners are performing (Wang et al., 2019).
Empirical research has grown in this area. Hu (2024) evaluated the readability of teaching content generated by ChatGPT for Business English courses, using Flesch, Dale-Chall and McAlpine EFLAW formulas. Nguyen (2025) conducted a questionnaire-based study with 42 undergraduate legal English students at Ho Chi Minh City University of Law, investigating the challenges students face when incorporating AI tools into legal writing. The findings revealed three primary challenges: overreliance on AI tools, educational ethics concerns regarding plagiarism, and reduced creativity, providing important pedagogical implications for guiding students to use AI tools responsibly and effectively in legal English contexts.
2.2. Current Research on Business English Instruction
Business English is a major branch of English for Specific Purposes (ESP). It helps prepare globally-competent professionals who can bring international perspectives to their work. The IBL course combines fields connected to ESP. It asks instructors and learners to reach high standards in legal knowledge and English communication skills. The course covers several areas of law. These include contract law, sale of goods, maritime law and international commercial arbitration. The language used in the course requires learners to know specialized terms, such as Incoterms 2020 terms and CISG provisions, along with complex clause structures and formal writing conventions common in legal work.
Conventional approaches to IBL instruction have several limitations. They depend too much on Chinese-language textbooks and cases, so the amount of English input stays low. The English cases that teachers develop on their own differ greatly in quality. Teaching is usually delivered through lectures, and there is not much interactivity or individualized scaffolding for learners. Assessment work mostly relies on summative terminal examinations, which cannot record how students develop their learning over time.
The existing literature on Business English teaching sees technology integration as a workable response to these limitations. Scholars have noted that contextual authenticity in language learning materials affects how well knowledge transfers. When scenarios match the communication needs that learners will face in the workplace, they have higher chances to use the skills they gain in simulated environments for professional work. At present, AIGC technologies have not been studied systematically for what they can do in Business English IBL instruction. This area of instruction needs to address disciplinary knowledge, language proficiency and practical competence at the same time.
2.3. Theoretical Framework
This study is designed on two complementary theoretical perspectives, namely, the Production-Oriented Approach (POA) and Personalized Learning theory.
Wen (2015, 2018) developed the Production-Oriented Approach, and this framework was built especially for Chinese foreign language education. It connects language learning with practical use. The POA holds that language acquisition is driven by meaningful output tasks. These tasks allow people to get cognitively involved with both language form and meaning. In the IBL context, AIGC can generate case analyses, mock court exercises and simulated dialogues. These activities fit the output tasks defined in the POA. Students must bring together legal knowledge and linguistic resources to finish these tasks, which helps them process content more thoroughly, thus making learning last longer.
Personalized learning originates from Bloom’s (1984) mastery learning model, and contemporary adaptive learning research (Viberg et al., 2023) has expanded this framework. The core idea is that learning outcomes work best when instruction is adjusted to match how far each learner has progressed, the path they are taking through study, and their specific needs. AIGC systems can bring personalized learning to large groups of learners. Through intelligent diagnostic assessment and learning pathway analysis, these tools adapt to the needs of individual learners. They give simplified explanations to students who need extra help, they share advanced case materials with learners who progress quickly, and they deliver real-time corrective feedback that matches the specific errors each learner makes.
Combined together, these theoretical bases give a clear reason for the AIGC-enhanced instructional model proposed in this study. The POA works as the pedagogical mechanism, and the Personalized Learning theory lets individualized instruction work across large groups of learners.
3. Research Methodology
3.1. Research Design and Participant
This study used a quasi-experimental research design that included a non-equivalent control group. Participants were Business English undergraduates from Baoding University. Forty-eight students who took the IBL course in the second semester of the 2024-2025 academic year made up the experimental group. Fifty-five students who took the same course in the second semester of the 2023-2024 academic year made up the control group. To ensure instructional consistency and research validity, the same instructor, syllabus, textbook, contact hours, and grading policy were used for both cohorts. Besides, because the two cohorts did not overlap in time, researchers conducted a pre-test to confirm baseline equivalence.
3.2. Research Instruments
The study adopted a mixed-methods approach that combined quantitative and qualitative data collection tools.
First comes the quasi-experimental assessment. The research team evaluated student performance with parallel pre-test and post-test tools. The team gave these tests at the start and the end of the target semester. The pre-test checked basic English proficiency and limited basic legal English, with a maximum score of 100. The pre-test paper mainly consists of two parts. Part I covers grammar, vocabulary, reading comprehension, and writing, which are 60 points in total; Part II assesses legal terminology and legal text understanding, which are 40 points in total. The post-test has three parts. IBL knowledge comprehension takes up 40 points, case analysis competence takes up 35 points and legal English proficiency takes up 25 points. The total maximum score of the post-test is 100.
Second is the survey questionnaire. At the end of the course, the experimental group filled out the AIGC-Assisted IBL Teaching Effectiveness Survey. Each item is scored through a five-point Likert scale, where 1 stands for strongly disagree and 5 stands for strongly agree.
Finally, the research team collected data through semi-structured interviews and thematic analysis (Braun & Clarke, 2006). 15 students were recruited from the experimental group and two course instructors. The students were selected based on purposive sampling. Each individual interview ran between 30 and 60 minutes.
4. Research Implementation
4.1. Needs Analysis and Preparation
Prior to the formal instructional experiment, the research team ran a diagnostic pre-test to find the most obvious problems that came with conventional teaching methods. We collected responses from the experimental group, and the main concerns are listed below. 78.4% of students said legal English terminology was too hard to understand. 67.7% said they did not get enough English case materials to study. 61.3% said they rarely got chances to practice what they learned, and 54.8% were not happy with the current assessment rules. These results showed the four-pillar model (AIGC-assisted teaching, personalized language practice, virtual internship and multi-model assessment) fit our needs, and helped us work out the detailed design for rolling out AIGC tools.
4.2. Course Development and Platform Construction
In the development phase, work centered on building the required resource base and technical infrastructure for the AIGC-enhanced model.
The research team built the AIGC case repository and assembled a corpus of authentic English-language IBL materials. These include the official CISG English text, ICC arbitration case abstracts, provisions from English law, and anonymous cases drawn from the business operations of Baoding-based import-export enterprises. Members of the team used these materials to fine-tune AIGC models, so the models can generate English cases that fit teaching needs.
Next, the research team built the online learning environment. They developed an integrated online platform that has several core functions. The platform supports browsing and searching the AIGC case repository, and it also gives personalized learning path recommendations. There is an AI-annotated submission portal for case analysis assignments, and there are virtual moot court modules, and a learning analytics dashboard that shows real-time progress visualization. The team finished writing all lesson plans, and they prepared accompanying PowerPoint slides for all units on the teaching platform Xuexitong. Each unit includes an AIGC-generated scenario introduction, a glossary of key terminology, an English case text, discussion questions and extended reading materials.
To realize the four-pillar teaching model, this study leverages Doubao, Huiya-chat-mini, Deepseek-R1 and Coze as core large language models. These tools generate preview guides, situational legal cases, and personalized language training content for pre-class, in-class, and post-class stages. For the English case repository, AIGC models such as Seedance 2.0 produce high-quality cases covering CISG, Incoterms 2020, etc., which are verified by legal instructors. The Chaoxing Xuexitong platform provides the function of AI assistant-teacher, which acts as the “language partner” as well as “legal expert” for post-class practice. Virtual practice modules are built via AIGC, based on real business cases contributed by Baoding-based import-export enterprises, such as the Juli Group. In the multi-modal evaluation system, Chaoxing Xuexitong platform provides multi-source data (participation, performance) to generate comprehensive reports. All AI outputs are strictly reviewed by instructors before classroom use to ensure accuracy and guarantee no intellectual property rights issues.
The instructional experiment ran in the experimental group for two straight semesters. Integrated with AIGC technology, the instruction of the course goes through four phases. The first part is a pre-class autonomous study, which lets students use AIGC tools to preview terminology and related background. The second part is in-class interactive teaching, which has instructor-led lectures and AIGC case discussions to extend the lectures. The third part is post-class personalized practice, which requires students to finish AI conversation practice and complete case analysis assignments. The last part is virtual practice sessions.
The control group received conventional instruction. Teaching centered on lectures, with textbook case analyses added. Final summative exams were the main way to assess learning. During the study, four categories of data were gathered. They were pre-test scores and post-test scores, connected with survey responses, processual learning data and interview records.
5. Results and Analysis
5.1. Student Performance Data Analysis
5.1.1. Pre-Test Equivalence
To ensure the authenticity and validity of the test results, both tests, including both the pre-test and post-test, were graded by the same teacher in accordance with the standardized scoring rubric.
According to the pre-test result, the independent-samples t-test found no significant pre-test difference between the experimental group and the control group (t = −1.11, p = 0.27 > 0.05). This means the two groups had similar baseline IBL knowledge before the intervention started (Table 1).
Table 1. Pre-test result.
Group |
N |
Mean |
SD |
t |
p |
Sig. |
Experimental |
48 |
65.7 |
10.23 |
−1.11 |
0.27 |
n.s. |
Control |
55 |
67.9 |
9.87 |
- |
- |
- |
5.1.2. Post-Test Comparison
After one complete semester of instruction, the post-test results showed the experimental group had statistically significant advantages in all three performance dimensions (Table 2).
Table 2. Post-test comparison.
Dimension |
Experimental (n = 48, M ± SD) |
Control (n = 55, M ± SD) |
t |
p |
Cohen’s d |
IBL Knowledge (40 pts) |
32.8 ± 4.12 |
29.1 ± 4.56 |
4.33 |
<0.001 |
0.86 |
Case Analysis (35 pts) |
27.5 ± 3.87 |
23.2 ± 4.21 |
5.64 |
<0.001 |
1.06 |
Legal English (25 pts) |
16.6 ± 3.14 |
15.8 ± 3.62 |
1.19 |
0.236 |
0.24 |
Total (100 pts) |
76.9 ± 9.34 |
68.1 ± 10.18 |
4.72 |
<0.001 |
0.88 |
The experimental group scored an average of 8.8 points higher than the control group on the total test. The difference reached high statistical significance (t = 4.71, p < 0.001). The largest effect size came from the Case Analysis dimension (Cohen’s d = 1.06). This means the AIGC-enhanced model worked well when it came to helping students build the ability to apply laws.
5.2. Survey Results
At the end of the term, researchers sent a questionnaire to the experimental group. Participants rated each item on a five-point Likert scale (1 = strongly disagree, 5 = strongly agree). Items were grouped into four dimensions. They are perceived utility of AIGC tools with 8 items, satisfaction with learning outcomes with 8 items, course experience with 6 items and improvement suggestions as 2 final open questions. Table 3 presents the descriptive statistics for the first three questionnaire dimensions.
Table 3. Survey dimension figures.
Dimension |
Items |
M |
SD |
Agree/Strongly Agree (≥ 4) |
Perceived Utility of AIGC Tools |
8 |
4.21 |
0.67 |
85.0% |
Learning Outcome Satisfaction |
8 |
4.38 |
0.59 |
91.2% |
Course Experience |
6 |
4.29 |
0.72 |
87.5% |
Overall |
22 |
4.29 |
0.64 |
91.2% |
The highest-rated individual item was “AIGC tools helped me better understand legal English terminology” with an M score of 4.52. The next one was “AIGC-generated English cases were highly beneficial for my learning,” with an M score of 4.46. The third one was “AIGC conversation practice improved my legal English oral expression” with an M score of 4.28. Students saw that AIGC tools had practical use in this area.
When it comes to how satisfied students were with the course, 94% of people who answered the survey said they were generally relaxed with the AIGC-enhanced model, and 89% said they would keep using AIGC tools in future courses.
As for the final part, open questions for suggestions and improvement, students who offered suggestions to improve the course quality asked for more cases that fit local conditions, and these cases came from China’s own foreign trade practice. This request came from 68.3 percent of respondents.
5.3. Interview Findings
After the completion of the AIGC-enhanced teaching in the experimental class, the research team conducted an interview with eight students and two instructors.
5.3.1. Students’ Feedback
The questionnaire for students covers five parts: learning effectiveness, AIGC tool experience, virtual practice platform, and overall suggestions. The aim of this interview is to explore students’ progress, feedback, and improvement ideas to assess teaching effectiveness.
Theme 1: Language proficiency gains. Most interviewed students said their legal English fluency and accuracy got better after they kept talking with the AIGC conversation partner.
Theme 2: AI-enhanced case learning. Most students thought cases generated by AIGC were more vivid. They had clearer context than textbook cases, and the difficulty matched each student’s own level. Some students said the case bank let them feel what it was like to successfully analyze real legal problems in English, something they had never done before.
Theme 3: Virtual practice platform that motivates engagement. Students and teaching staff have been positive about the moot court module. Students found deep immersion in their assigned roles helped them want to learn more.
Theme 4: Personalized feedback helps students learn on their own. Many students noticed that AIGC tools give immediate personalized feedback. It connects three steps: finding errors, getting corrective input, and making progress. This process lets them keep going farther with independent study.
5.3.2. Instructor Feedback
The teachers’ interview covers instructional design, student performance, AIGC advantages and limitations, teacher‑AI collaboration, multi-modal evaluation, and future prospects, aiming to evaluate teaching practice and provide professional suggestions.
Instructor A said that “AIGC tools have substantially lightened my lesson preparation burden, especially the case repository, which has solved the longstanding problem of English case scarcity. The AI evaluation system also lets me understand each student’s learning progress more fully.” Instructor A pointed out, however, that “AIGC tools cannot fully replace the instructor, particularly in value judgment on legal issues and the cultivation of professional ethics. In these areas, human instructors still fill a role that cannot be taken over.” Instructor B said that “the multi-modal assessment system has made evaluation more equitable and comprehensive. Previously, the final examination determined most of the grade, and many students could score well by cramming right before the test. After more focus is put on assessments that track learning over the whole process, students now approach their learning and practice more seriously throughout the entire semester.”
6. Discussion and Suggestions
6.1. Effectiveness of the AIGC-Enhanced Instructional Model
This study finds clear evidence that the four-pillar AIGC-enhanced model helped improve teaching effectiveness in the Business English IBL course. The experimental group performed better in every dimension we measured. This means the differences we saw have meaningful implications for educational practice, and they also hold up to statistical testing.
It is found that the share of students scoring below 60 points has fallen. It went from 12.1 percent in the control group to 3.2 percent in the experimental group, and this change is especially notable. Fewer low-scoring students in the distribution of grades shows that personalized supports built into AIGC tools can help students who have trouble keeping up with traditional classroom learning.
6.2. Industry-Education Integration through Virtual Platforms
The virtual practice platform includes real cases from foreign trade enterprises based in Baoding. One of these enterprises is the Juli Group. This design keeps practice content connected to actual business contexts. Interview results show students are happy with learning scenarios tied to local development. The outcome shows that industry-education cooperation can help improve courses that use AIGC.
6.3. Advantages of the Multi-Modal Assessment System
Traditional IBL courses typically use summative final examinations as the main way to assess student learning. This common practice cannot fully capture how professional competency develops across multiple areas. The multi-modal assessment framework used in this study fixed this problem. It spread evaluation across four distinct dimensions. It also used AIGC tools to give immediate and individualized feedback on every student submission. AI-generated evaluation comments shorten the “learning-feedback” cycle. This change seems to help students correct errors faster and refine their conceptual understanding. This effect matches findings recorded in existing adaptive learning research (Bloom, 1984).
6.4. Limitations
This study has several limitations that need to be noted. First, even though the sample size was large enough to find meaningful results, the total number of participants was not very large. Forty-eight participants were in the experimental group, and fifty-five were in the control group. More studies with larger groups of students from different backgrounds need to be done before results can be widely applied. Second, because the IBL course-teaching only took one semester in a whole academic year, though the whole research lasted for two years, the practical teaching ran for only two semesters, one semester employing traditional teaching for the control group, while another semester for the experimental group. It can only tell us about effects that last over the medium term. It cannot answer questions about how long knowledge will be kept, or how this affects students’ future career choices. Third, the study found that content generated by AIGC sometimes had inaccurate information or mismatches with the context. Teachers still have to check the content before using it. This reliance on teacher review affects how widely the tool can be used, and it adds to teachers’ workload. AI tools can greatly increase the number of language practice opportunities available to students and help these opportunities reach more learners. The effectiveness of these tools still depends on how teachers build guidance around them and how they fit the tools into established teaching frameworks (Libunao & Morales, 2025).
6.5. Suggestions
Based on these findings, we offer the following recommendations to educators and administrators who plan to bring AIGC into ESP courses. First, AIGC tools can work alongside qualified instructors to support teaching; they cannot take the place of these instructors. Instructors still have to lead students to think through ethical issues, shape students’ professional values, and help students conduct detailed legal analysis. This work cannot be done by tools. Second, cases made or checked by AIGC tools need to fit the specific industry that matches the career students plan to enter after graduation. Schools can establish partnerships with local businesses and legal service providers to keep course content aligned with actual work needs. Third, when the teacher adds AIGC to his teaching, he needs to adjust it over time. Besides, he should collect systematic information on how students perform and gather their feedback.
7. Conclusion
Following the strategic deployment of national education digitalization in China, this study integrates the technology of AIGC into the whole process of the International Business Law course teaching. It builds up the four-pillar instructional model and meanwhile verifies the effectiveness of this system. As the result of this study demonstrates, the technology of AIGC is not only an effective tool to solve the current difficulties confronting the teaching of International Business Law, but also provides valuable practical aid and momentum to the digital transformation of higher foreign language education. With the continuous development and upgrading of artificial intelligence, AIGC technology will definitely be used in more scenes and fields.
Acknowledgements
This paper is under the project of AIGC-Enhanced Teaching of International Business Law Course in Business English Majors: Innovation and Practice (2024YYJG056) sponsored by the 2024 Hebei Provincial College English Teaching Reform Research and Practice Projects. I am deeply indebted to all the team members, who offered invaluable advice and comments.