1. Introduction

In recent years, artificial intelligence (AI) has permeated various aspects of human life, particularly in the educational domain. Specifically, in the context of English language learning, the integration of AI technologies is progressively transforming conventional pedagogical approaches, leading to significant paradigm shifts in instructional methodologies [1].

Meanwhile, the post-pandemic era has witnessed the most notable shift in the accelerated adoption and implementation of online education and AI-powered tools. Particularly for students who enrolled in universities in 2023, there has been a marked increase in both reliance on and habitual use of digital learning platforms. Previous research indicates that a significant number of students reported a higher frequency of utilizing AI-based tools for English language acquisition after commencing their tertiary education,teachers unanimously identify this trend as a pressing pedagogical challenge, emphasizing its potential to undermine academic integrity and independent critical thinking [2].

Although AI has brought negative effects on college students who over dependence on it, it is also progressively developing customized and optimized educational approaches, which demonstrates measurable improvements in knowledge retention and skill acquisition compared to traditional teaching models [3]. Reading is an essential part of college English education, especially for English major students, however, under the traditional task-based teaching model, English reading tasks tend to be monotonous, lacking diversity and engagement. Moreover, the uniform task design fails to provide individualized guidance for each student. The evaluation system predominantly relies on teacher assessment, employing a singular criterion that inadequately reflects students' learning outcomes and progress in a comprehensive and objective manner [4]. However, the emergence of artificial intelligence technologies offers innovative approaches to English education instruction [5].

Given that female students are in the majority within English majors, and considering the feminist identity development among elite young women in Chinese university environments in recent years, as well as the fact that the second academic year is the period when English majors at this university begin to encounter a substantial volume of academic English reading tasks, this study specifically focuses on second-year female English major students. Additionally, the educational development in the coastal areas of Guangdong Province is at the forefront of China's educational innovation [6]. This study interviewed five female English majors in Guangdong, China, to explore their dependence on AI-assisted reading and potential improvements in college English education. The research aimed to identify ways to eliminate AI reliance and compare AI assistance with teachers, providing suggestions for future research on AI's optimal use.

2. Methodology

In this study, field work was conducted between March and April in 2025 at University S in China’s coastal city, Shantou. S University has many great English teaching resources for second language learners, the in-depth cooperation policy between S University and some foreign universities provides English learners and English major students with a rich foreign language learning environment and opportunities for studying abroad.

Five female students in S University ‘s English major were picked for interview, they are all sophomores who have just come into contact with academic writing courses and academic English texts. In the interview, participants were all asked to answer the questions about their point of view about AI assist English reading, the comparison between human teachers and AI and their way to resist AI reliance. The data analysis started during the fieldwork period, and was completed through repeated observation and summary of the transcribed text of the interview.

3. Discussion

In the interview, five female college students in English major have answered 10 questions from three main aspects:

1. The reasons for choosing AI assist English reading

2. Comparison of the assistance between AI and college English teachers

3.College students’ reliance on AI

3.1. The reasons for choosing AI assist English reading

During the interviews, five female college students shared their profound insights on this topic through interviews. When asked the reasons of choosing AI-assisted English reading, thier answers were surprisingly consistent, all of them pointing out that the immediate cause of using AI to assist English reading is the long, boring or difficult to understand of the reading context, one interviewee especially pointed out that the academic papers and cryptic poetry are particularly difficult to read:

“Ugh, it’s way too long and boring—plus I just don’t get it.” (Wanjing He)

“Whenever I see a huge wall of text, it feels super overwhelming and hard to digest, so I figured AI could help me break it down.” (Xintong Ji)

“First off, the articles are just way too lengthy, and there are way too many unfamiliar words. I just wanted a quick fix to get through the reading.” (Dan Ye)

“I turn to AI when I’m totally lost or when I’m in a time crunch—like, I’ll just toss the document at it and ask for a summary or outline.” (Yuetong Wang)

“Honestly, it depends on the text. If it’s something dense, like academic papers or cryptic poetry, that’s when I’ll bring in AI to help me out.” (Qiuju Li)

Five English major students in Guangdong faced high-burden assignments, leading them to seek AI help. They revealed that AI could provide abilities and knowledge they couldn't acquire in college classes, highlighting the need for more support.

Wanjing He pointed out that: “AI helps me quickly extract key points from a text, which makes it easier for me to dive deeper into understanding the material afterward.” AI helps her quickly grasp the key messages of the text. This ability is highly required in the TEM4 test, which English major students in China must pass to get its certification during college. However, there’s seldom instruction of this kind of ability in college classes, as college teachers have their own teaching schedules to follow. The emergence of AI has lend a helping hand to students like Wanjing He.

Qiuju Li said: “I mostly use AI for tricky texts, like when I’m stuck on dense academic papers. Also, my grammar’s pretty weak, so I rely on AI to break down sentence structures.” In China, students often lack grammar practice and knowledge before entering college due to various factors such as local education policies, economics, and the Gaokao exam. High school teachers often assume students have a solid foundation in grammar, neglecting it in class, leading to gaps in knowledge when entering college. College courses rarely provide remedial grammar lessons, forcing students to study independently. The emergence of AI has made self-study more convenient, but students may turn to the internet for help when facing problems in English study. Understanding students' behaviors towards AI and human teachers is crucial.

3.2. Comparison of the assistance between AI and college English teachers

As is widely acknowledged, Artificial intelligence (AI) enhances college English teaching by making it more interactive and practical. It supports modern intelligent education, improves teaching quality, and makes learning more interesting. It also enriches students’ free time and encourages creative thinking in English learning, helping students shift from passive to active learning. However, AI can never replace the emotional connection between teachers and students. Instead, in the age of AI-powered education, college English teaching is no longer just about the interaction between teachers and students. In the era of AI-assisted education, English instruction is no longer limited to the traditional interaction between teachers and students—AI has become a new and important participant in the teaching process. As a result, three key agents now shape the classroom dynamic: the teacher, AI, and the students [7].

Previous studies suggest that AI like ChatGPT can serve as helpful learning companions, supporting learners in exploring English learning activities in creative and effective ways. These studies also acknowledge the risks associated with AI, including concerns about cheating and fair assessment. They emphasize that rather than avoiding AI, students should learn to use it wisely, with teachers playing a guiding role in helping them leverage AI for personalized, extracurricular learning [8].

In the interviews, several questions about comparison between AI and college English teachers were asked, when asking interviewees to compare AI and college English teachers in the aspect of reading comprehension, Dan Ye answered: “ AI explains reading materials more systematically than university teachers. Currently, professors usually only address specific problems when teaching reading, but AI helps me understand the text more comprehensively.” Dan Ye’s point of view reveals that some college English learning courses are still keeping the traditional teaching mode, which is characterized by serious exam-oriented thoughts, and the neglect of improving application ability, which can lead to college students’ rigidity of thinking [9].

Yuetong Wang provided a different point of view: “I think college teachers focus more on sharing ideas, while AI provides more factual knowledge. With teachers, you can keep asking questions, and they’ll give clear, relevant answers—unlike AI, which sometimes responds inaccurately or even gives misleading information that needs fact-checking. However, AI saves time, and you can gain more knowledge from it in one session compared to a single lecture.” Her view reveals that AI cannot completely replace teachers in college. Additionally, AI can sometimes be misleading. Therefore, it is essential for college students to cultivate critical thinking skills.

Despite these limitations, AI still plays a useful role in students’ pre-class preparation and after-class review. As Xintong Ji pointed out: “Teachers aren’t extremely helpful—most of the time, I rely on reading the text myself. University instructors don’t seem to assist much with English reading because they don’t explain word by word. I have to read and grasp the advanced meaning first; otherwise, I won’t understand anything in class.” After entering college, teachers have their own research tasks to complete, so they won’t spend all of their time focusing on step-by-step teaching as they did in high school, just as she said. This also means that students need to handle their studies mostly in the preview and review sections by themselves, and AI can now fuel their power to head into this process.

3.3. College students’ AI reliance

AI’s birth has raised concerns among education area, people are holding different point of view about college students’ attachment with AI. New York City’s education department has blocked access to ChatGPT on school devices and networks due to concerns about its impact on learning, content safety, and accuracy. Officials worry the AI chatbot, capable of producing high-quality essays, might discourage critical thinking and encourage cheating. While the tool provides quick answers, it doesn’t help students develop problem-solving skills. However, schools can request access if they want to study the technology. This decision by the largest U.S. school system may influence other districts dealing with AI in education [10]. There’s also educators like Xi Wang and Ke Yan, who actively explore new methods and possibilities of integrating AI technology into teaching, they integrate POA into College English reading teaching and work out good results [9]. The study investigates students' views on AI reliance in reading and their suggestions for reducing it. Both interviewees admit to AI reliance and seek help when reading difficult materials. To eliminate AI dependence, students suggest creating a learning environment that encourages independent thinking and staying away from electronic devices. Yuetong Wang advocates for self-restraint and allowing students to work through problems on their own. Qiuju Li suggests understanding AI as a tool, not the foundation of learning, and setting clear boundaries when using it. This approach helps students return to rational thinking and avoids relying solely on AI. Just as Wanjing He and Qiuju Li said, people with poor self restriction should pay attention to cultivate their independence thinking skill, and force themselves to read the material by their own at first instead of seeking help from AI immediately, and keep themselves in an environment that helpful for deep thinking is also a good choice. Attitude likes “Allow yourself to spend more time working through the problem on your own.” can also become a positive and effective psychological hint when students are trapped by the difficulties in reading.

4. Conclusion

In today’s academic environment, the widespread use of AI among college students has brought considerable convenience, particularly in navigating and comprehending lengthy or abstract materials such as research papers and poems. However, AI can’t replace human teachers. Instead, it can become a positive and helpful tool applied to college English reading classes and make up for human teachers’ shortcomings. By leveraging AI’s advantages, the deficiencies in college English education can be better recognized, as the interview results in this paper reveal several points: less systematic explanation in reading, limited classroom interaction depth, less focus on detailed text analysis, and traditional exam-focused teaching methods that neglect the improvement of practical application skills. These aspects should be considered in further improvements to college English teaching. The study also highlights students’ own ways to reduce AI reliance. Regarding the methods for interviewees to decrease dependence on AI, they can be summarized into two aspects: external support and personal mindset development. The most crucial factors are cultivating critical thinking skills and the ability to evaluate and selectively use AI-generated content. Some limitations should be acknowledged, such as the involvement of only five participants and the lack of perspectives from teachers and other educators. Moreover, the subjectivity of the interviewees’ answers and their limitations in levels of knowledge and experience with AI technologies should also be taken into consideration when interpreting the findings.

Ultimately, this paper aims to inspire educators in applying AI to future education improvement in college English reading and formulating new plans about reducing AI reliance in college English reading by recording students’ reflections on AI dependency.