





































  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

167 
 

THE USE OF HUME AI IN SPEAKING ASSESSMENTS: 
TEACHERS’ PERCEPTIONS IN SENIOR HIGH SCHOOL 

CONTEXTS 
Nisa Fahria Nasution1, Diah Safithri Armin2 

1,2Universitas Islam Negeri Sumatra Utara, Medan, Indonesia 
 

nisa0304212059@uinsu.ac.id 

  

ABSTRACT 

This research aims to investigate the perception of teachers about the use of Hume AI for 
assessing students’ speaking skills in senior high school. English teachers (n=4) from public 
senior high schools in North Sumatra participated in this study. This research employed a 
descriptive qualitative method with semi-structured interviews and field note observations. 
The data was examined by using thematic analysis. The result indicated that Hume AI has a 
positive impact on students’ enthusiasm and interest. However, several challenges hinder 
the optimal use of Hume AI in speaking assessments, particularly in spontaneous speaking, 
including their low motivation to learn English due to socio-economic factors, and 
insufficient access to technological infrastructure within schools. Despite these obstacles, 
teachers acknowledged the potential of Hume AI to enhance speaking assessment and 
recommended its integration as both a learning and assessment tool. These findings offer 
insightful information for teachers, policymakers, and AI developers looking to implement 
AI-based tools in language learning and assessment practices.  

Keywords: Hume AI, Learning Media, Speaking Assessment, Teachers’ Perceptions. 

 
INTRODUCTION 

Teachers hold a crucial position in the realm of English learning, teaching exactly in 
designing, executing, and evaluating (Hattie, 2012; Mundy et al., 2008). One of the most 
crucial functions of a teacher in a language classroom is to serve as an organizer, as it 
requires organizing both the students and the variety of activities involved in language 
learning (Naibaho, 2019). The English teacher was an organizer who created lesson plans, 
organized the syllabus, and oversaw the teaching and learning process (Budiharto & Affandi, 
2018). As an executor, in addition to overseeing all classroom activities, the English teacher 
is in charge of the class. In this case, English teachers will transfer the knowledge they have 
to students. In general, an English teacher is in charge of assessing student progress 
regarding language proficiency as well as their areas of strength and weakness (Poudel, 
2022).  

In the conventional learning and teaching process, teachers are the central source of 
information, in which the teachers explain and students only listen, which is nowadays being 
reduced (Anggreini et al., 2023). Recently, the system began to change to be a learning 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689
mailto:nisa0304212059@uinsu.ac.id


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

168 
 

media-based method. Information-distributing media that are utilized to facilitate the 
learning process are referred to as learning media. Learning media constitute an integral and 
crucial component in determining the effectiveness of information delivery, particularly as 
tools that support student assessment. Their appropriate use ensures that the intended 
learning outcomes, as outlined in the instructional design, can be successfully and optimally 
achieved (Mutia et al., 2020).  

The efficiency of the learning process in the classroom can be assessed by the extent of 
students' interest and active engagement in the instructional activities facilitated by the 
teacher. In order to ensure that students remain motivated throughout the learning process, 
it is the responsibility and duty of teachers to consistently maintain and enhance their 
students' motivation to learn. Teachers must actively seek effective strategies to boost 
students' enthusiasm for learning, address declines in motivation, encourage self-directed 
learning, and cultivate a sustained sense of motivation within the learners themselves 
(Dwijuliani et al., 2021). Therefore, it is essential to employ specific techniques and 
strategies that support the effectiveness of the teaching and learning process. 

Technology-based interactive media can significantly enhance the learning process, as 
technology provides a hardware-oriented approach that facilitates the implementation of 
educational activities. The use of technology-based interactive media has the potential to 
actively engage all students in the learning experience. This involves the use of various 
instructional tools such as teaching machines, films, slides, simulators, overhead projectors, 
and videotape recorders (Kustyarini et al., 2020). 

In the face of global challenges such as the climate crisis, security threats, and the 
COVID-19 pandemic, societies are compelled to explore and implement innovative, forward-
thinking solutions to ensure resilience and sustainability (Stenbom & Geijer, 2025). Digital 
transformation, driven by technological innovation, provides a strategic avenue for re-
envisioning and strengthening resilience across multiple sectors, particularly within the 
domain of education (Lasi et al., 2014; Vial, 2021). This transformation, characterized by the 
intricate convergence of cultural, political, and economic dynamics, fundamentally 
reconfigures human-technology interaction and engenders far-reaching implications for 
educational systems and theoretical paradigms, thereby necessitating a comprehensive 
reexamination of pedagogical practices and learning methodologies (Berry & Dieter, 2015; 
Gorecky et al., 2014). 

Likewise, Indonesia's education sector is witnessing a significant increase in the 
integration of information technology. This transition reflects the sector's adaptive response 
to the demands and challenges posed by an increasingly dynamic and fast-evolving era 
(Widodo et al., 2021). Ongoing technological advancements continually contribute to the 
increased efficiency and practicality of various facets of human life. Among the most recent 
developments is the emergence of artificial intelligence (AI), which has garnered substantial 
scholarly and societal attention for its potential to emulate human cognition and behavior 
(Fitria, 2025). 

The development of tools for generative artificial intelligence (GenAI), such as 
ChatGPT, has raised challenging questions about the nature of teaching, learning, and 
assessment in schools and universities (Tang et al., 2024). GenAI is expected to revolutionize 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

169 
 

and change how we learn, think, and work by producing original material quickly. It will 
quickly make everyone who does not comprehend it feel left behind. The usage of AI software 
and robots as assistants, tutors, and peer learning facilitators in classrooms around the 
world is not surprising given the growing integration of AI technology into education 
(Vasagar, 2017). Artificial intelligence (AI) has been utilized in education for more than three 
decades, with applications such as predicting student performance, personalizing learning 
experiences, and automating assignment grading. In recent years, the integration of AI in 
education has become even more prominent, particularly following the release of advanced 
AI tools such as ChatGPT (Zawacki-Richter et al., 2019; Zhu et al., 2019). For the effective 
integration of artificial intelligence (AI) into educational practice, it is essential for educators 
to systematically evaluate methods for maximizing its positive impact on learning outcomes, 
while concurrently addressing potential risks and mitigating any adverse effects (Kim, 
2021). 

In Indonesia, the integration of artificial intelligence into educational media remains 
limited, primarily due to the unequal distribution of adequate technological infrastructure 
across schools (Helmiatin et al., 2024). Moreover, the utilization of artificial intelligence in 
Indonesia remains highly significant, particularly for educators and technology developers 
who advocate for the integration of AI-driven learning strategies within educational contexts 
(Jo, 2024). 

In the field of education, assessing students’ activities of various types is an integral 
part of teachers’ workload (Alsalem, 2024). As asserted by (Beaumont et al., 2011), 
assessment has long been recognized as a crucial element that significantly contributes to 
both teaching and learning processes. Anton, within the field of ELT (Anton, 2009), argue 
that assessment plays a vital role in fostering the learning process and can effectively 
support learners in enhancing their academic development (Safdari & Fathi, 2020). 
According to Permendikbud No. 104 of 2014, Tests and attitude scales are examples of 
assessment instruments, which are tools used to evaluate student learning results.  

Speaking is thought to be the most difficult of the four language skills (speaking, 
listening, reading, and writing) for junior high school teachers to evaluate (Navidinia et al., 
2019; Rahmawati & Ertin, 2014; Zaim, 2017). Speaking ability examinations are used to 
assess students' speaking proficiency. The assessed aspects common for the speaking test 
are addressed to fluency, accuracy, pronunciation, grammar, vocabulary, and gesture 
(Brown, 2007). Learners can be considered to possess strong speaking skills if they 
demonstrate proficiency in the following categories (Aprianoto & Haerazi, 2019).  
Nevertheless, empirical research focusing on the development and assessment of learners’ 
speaking proficiency remains relatively scarce (Vercellotti, 2015).  

There are extensive studies exploring the use of various media to support speaking 
assessments; this area remains a subject of ongoing investigation. As Hung and Huang (Hung 
& Huang, 2015), examined the application of Web 2.0 tools, specifically blogs, as instruments 
for assessing and evaluating EFL learners' speaking performance. Their study revealed that 
blogs significantly contribute to the development of students' oral presentation skills and 
foster a positive learning experience. 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

170 
 

Nevertheless, the focus of research should extend beyond assessment tools to include 
the underlying frameworks and assessment models that inform the evaluation of speaking 
proficiency. In this regard, Safdari and Fathi (Safdari & Fathi, 2020) highlighted Dynamic 
Assessment (DA) as a process-oriented, alternative assessment approach that promotes 
learners' active engagement and autonomy in the learning process (Crick & Yu, 2008). DA 
has a significant impact on speaking accuracy of the participant; however, for speaking 
fluency, DA did not significantly improve.  

Based on some of the previous studies above, there are still very few studies that 
discuss AI as a tool in speaking assessment. Therefore, this research was conducted to find 
out how teachers perceive an AI used to assess students' speaking skills. Specifically, the 
focus of this research are: 1) how teachers perceive the Hume AI as a tool to assess students’ 
speaking skills in senior high school? 2) what are the challenges faced by teachers while 
using this tool for assessing speaking skill? Top of FormBottom of Form 

The findings of this study offer valuable insights for readers, particularly teachers, 
regarding the potential use of artificial intelligence, specifically Hume AI, as a tool for 
assessing students' speaking skills. The integration of such technology is expected to provide 
practical benefits for teachers by facilitating a more efficient and objective assessment 
process. Furthermore, this study contributes to the existing body of literature and may serve 
as a reference for future researchers interested in exploring the application of Hume AI in 
educational assessment. In addition, the findings may serve as constructive input for AI 
developers, encouraging the development of applications that function not only as learning 
media but also as reliable assessment instruments. 

METHODS 

This study employed descriptive qualitative research to investigate the perspective of 
teachers while using the Hume AI to assess students’ speaking skills and employed 
phenomenology, which is specialized to emphasize the lived experiences of participants and 
is frequently based on in-depth interviews (Patton, 2002). This study used a qualitative 
method due to the desire to learn more about instructor perspectives and gain a deeper 
understanding of all of them.  

For the study, four English teachers in senior high school from four public schools in 
North Sumatra were chosen. The participants in this study were English teachers with over 
ten years of teaching experience. These teachers were selected based on their extensive 
professional background and their considerable experience in teaching a diverse range of 
students. All participants in this study were female teachers. Furthermore, these teachers 
were identified as individuals who are well-regarded by their students, as reflected in their 
positive rapport and favorable reputation within the school community. This intentional 
decision is warranted since the qualitative case study's sample selection procedure 
emphasizes the depth of the participants' comprehension of the problem rather than the 
population's representativeness.  

The data were collected by classroom observations and semi-structured interviews. To 
find out how teachers utilize Hume AI to evaluate students' speaking abilities and whether 
Hume AI can evaluate speaking skills in general, non-participant, less structured 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

171 
 

observations of each teacher and some students were carried out by field notes. The semi-
structured interview has the following section: 1) perception of teachers related to the use 
of Hume AI to assess speaking skill 2) challenges that are faced by teachers while using Hume 
AI. These broad topics were considered to cover the teachers’ holistic experience and 
perspective. 

Qualitative content analysis was employed to examine the information gathered from 
semi-structured interviews and classroom observations.  In this investigation, four stages of 
the analytical method were used. The verbatim transcription of the data was the first step in 
the data analysis process, which was conducted with the simultaneous gathering and 
interpretation of the qualitative data in mind. The researchers manually entered the 
transcriptions into word processing papers. All of the recordings were turned into texts 
because the qualitative element was the main emphasis. In order to become familiar with 
the facts, the texts were then read and reread in their entirety. The first step in the data 
reduction and interpretation process was text coding. The texts were divided into sections 
during the first coding process, and the parts were labelled, frequently in the participants' 
own words. Hard copy printouts of the texts were utilized for the first coding, and the 
margins were labelled. Higher-order themes were then created by grouping the original 
codes that shared concepts.  

RESULTS 

Teachers’ perceptions of students’ enthusiasm and interest in using Hume AI  

Respondents were asked to share their response on students’ enthusiasm and interest 
when using the Hume AI for assessment. The findings showed that teachers observed all 
students displayed heightened enthusiasm when first introduced to Hume AI as a tool for 
assessing their speaking skills. The novelty of technology like this AI seemed to capture their 
interest, leading to a more engaged learning environment. In the school, this AI was first 
implemented in English language learning as media for assessing speaking skill, but most 
students seem to like and be very interested in this AI. As one teacher noted,  
“So, if we continue to use this, I think it will grow because this way these students will be 
enthusiastic, especially when they first hear it”. 
Another teacher added:  
 I think it is the best one. I see they like it very much, and they are definitely motivated. I think 
it can improve their interest, their ability, and they will try to focus.  

The word ‘definitely motivated’ highlights the positive impact of using Hume AI on 
student motivation. The teacher indicated that the integration of technology in the classroom 
could enhance not only the interest and enthusiasm of students but also students' motivation 
to participate in speaking activities.  

In addition, findings from student observations support the respondents' statements, 
indicating that when Hume AI was first introduced and practiced in class, students expressed 
positive emotions, such as smiling and showing enthusiasm. The students listened 
attentively and in an orderly manner to the instructions on how to use Hume AI. They also 
actively asked questions when they encountered difficulties in understanding how to 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

172 
 

operate the application. Furthermore, during the practice session, many students showed 
interest and were eager to directly communicate with Hume AI. 

The reason why students are motivated and enthusiastic because the current 
generation, who are mostly high school students, are very exposed to technological 
developments. Therefore, they are more enthusiastic and excited when this AI is 
implemented in their English assessment. Like one teacher stated: 
“They look motivated because I think they are not far from items or technologies like this. Isn’t 
it?” 

The majority of senior high school students were able to engage with and benefit from 
English learning through the use of Hume AI. They also demonstrated ease in understanding 
the instructions for operating the AI during the speaking assessment conducted by the 
teacher. 

Teachers perceived the advantages of Hume AI for assessing speaking skill  

When participants were asked to say their opinion regarding the advantage of Hume 
AI, especially as a media for supporting speaking assessment, the majority of teachers 
reported that they actually seldom integrate technology into assessment practices. 
Technology is predominantly utilized for instructional purposes, such as delivering learning 
materials through presentations or videos. In contrast, assessments are generally conducted 
using conventional manual methods, as one teacher stated:  
“……., then the assessment is straightforward. Directly assessed. Manual.” 

In conventional manual methods, speaking assessments are typically conducted 
through dialogues or conversations based on written texts, which they subsequently 
memorize. Similarly, some teachers organize activities such as drama performances or 
debates; however, these tasks also rely on written scripts that students are expected to 
memorize in advance. Overall, the majority of speaking assessments conducted in these 
contexts require students to produce written texts, commit them to memory, and then 
present them orally in front of the class. 
“Yes, written and then read or practiced by the students in front of the class”. 

Teachers believe that measuring students' speaking abilities can be effectively 
conducted through practical speaking exercises. However, on average, teachers do the 
assessment manually so that when this AI was first implemented at school, especially in 
assessment, teachers seemed enthusiastic because they thought this AI could measure 
students' speaking skills. As the teacher stated: 
“If you measure students' abilities, it can be measured because Hume AI is directly practical, the 
AI directly talks to students.” 

Based on the observation of students who participated in the assessment using Hume 
AI, teachers appeared satisfied, as the use of this AI encouraged students to actively engage 
their thinking in order to understand and formulate appropriate responses during the 
assessment. Furthermore, the assessment process conducted by the teacher proceeded 
smoothly. 

The direct interaction, where students articulate their thoughts, allows for immediate 
assessment of their fluency, pronunciation, and overall speaking competence. Students will 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

173 
 

also be more focused on listening and answering questions and also train to be quick and 
responsive in giving and answering questions without preparing and memorizing the 
answer in advance. This suggests that real-time speaking activities are crucial for evaluating 
students' oral skills. 

On the other hand, teachers emphasize the importance of aligning AI assessments with 
the existing curriculum and syllabus. The expectation is that by the time students graduate 
from senior high school, they should be able to speak English fluently. Teachers stated: 
“It means that in senior high school, graduate from senior high school, students are expected to 
be able to speak English fluently. so that is an indicator of students who have learned English 
in senior high school.”   

Since one of the requirements for passing English learning in senior high school is the 
ability to communicate fluently in English, Hume AI can be utilized as a learning tool, 
particularly as an assessment instrument for English speaking skills. If students can speak 
English fluently with this AI whose background language is native speakers, students are 
expected to be able to face the world of work and college. As one teacher stated: 
“……After graduating from senior high school, students can be speaking directly to apply for  
jobs in English.”  

That is curricular goals, ensuring that students are adequately prepared for real-world 
applications of their language skills, such as job interviews. 

Teachers’ perceptions of challenges faced by teachers while using Hume AI for assessing 
speaking 

Participants were asked to share their insight on challenges faced by ELT while using 
Hume AI. The results indicate that one of the primary challenges encountered during the 
implementation of Hume AI for speaking assessment is students limited foundational 
knowledge of English-speaking skills. This was particularly evident during classroom 
observations, where students frequently exhibited confusion when responding to questions 
posed by the AI. It was also observed that students often sought clarification from the teacher 
regarding the meaning of the AI's questions, which disrupted the assessment process and 
hindered its effectiveness. Consequently, teachers were required to divide their attention 
between assessing students' speaking performance and translating or explaining the AI's 
prompts. In many cases, students who struggled to comprehend the AI's instructions 
remained silent or abandoned the assessment before it was completed. As one teacher 
stated: 
“………yes, that's how the students are here. They really don't know anything at all.”  

Students’ knowledge was only up to writing in a structured way, but to speak and 
interact directly using English, they had a bit of difficulty. There are some students who can 
speak well using the Hume AI, but that is only at the beginning; in the middle, many students 
do not know and understand what is being asked. Besides that, during the assessment 
process, many students are limited in vocabulary mastery. a lot of unfamiliar vocabulary they 
hear from the AI, so the students do not catch the meaning of the questions given. 

The reason why students do not have a strong English base is that students lack 
motivation and interest in learning English. 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

174 
 

“On average students show a lack of interest in learning, particularly in speaking English. We 
have given motivation, indeed lack of interest in learning, lack of persistence, especially in 
English. So, it's difficult if you make a media-based strategy for improving English listening 
skills, because students tend to lose focus and become disengaged from the beginning.” 

The loss of motivation and lack of interest of students is influenced by the economic 
situation of the family. The average student treated by AI has a middle to lower economy. 
Thus, making students not interested in learning, especially English, because they think 
learning English will not be useful for them. 
“Here the economy is lower middle class; most of their parents are fishermen. That means they 
have no interest in English. And also, they don't think learning English is important to them.” 

Another reason why students do not have a strong English base is because students do 
not take courses outside school. Most students are unable to pay for tutoring outside of 
school. As one teacher stated, ‘That's why it's difficult for students to take lessons outside. 
It's difficult to eat every day’. Several students who can communicate with this AI have a 
background and are currently taking English lessons outside of school. 
“………….at least from taking a course, it will help them to explore this lesson using this AI.” 

In addition, the challenge faced by teachers is insufficient facilities. Limited facilities 
are an obstacle in developing teaching methods. In English language learning, facilities are 
an important aspect to support the quality of learning. The 4 main aspects of English 
(writing, reading, listening, and speaking) mostly require adequate facilities so that students 
can easily practice. But some schools have limitations in having these tools or media. 
Because the school facilities are also less supportive, we are also delayed too. For facilities we 
are very much looking forward to media tools to help with English as a speaking tool. 

Based on observations in several schools, the facilities that support English learning 
are still very limited. For example, in some schools, there is only one projector available, 
which forces the teachers to take turns using the device. Regarding speakers, only a few 
schools provide high-quality speakers, while others do not have speakers at all.  

Due to the lack of facilities, teachers' ideas and plans to make lessons fun and 
interactive must always fail, especially during speaking lessons, which are always postponed 
later and later. This causes teachers to finally rarely use technology and use manual methods 
to keep lessons active and exciting. As one teacher said, 'Actually in this school, most of the 
teachers teach manually.’ 

Limitations such as the internet network and tools that support speaking assessments, 
such as speakers and projectors, are the weaknesses of this AI implemented in schools. Like 
teachers stated: 
“If it can be held, one of them is the internet network here. There is no network at school; besides 
that, there are no radio speakers here.” 

DISCUSSION 

This study set out to explore how teachers perceive Hume AI as a tool for assessing 
speaking skills among senior high school and the challenges faced by teachers while using 
this AI for assessing speaking. The findings revealed that the heightened enthusiasm and 
motivation demonstrated by students upon the introduction of Hume AI aligns with research 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

175 
 

highlighting the positive impact of technological innovation on student engagement (Fitria, 
2025; Vasagar, 2017). The novelty of AI-based assessment tools appeared to spark students' 
interest, supporting the notion that technology can enhance learner motivation and 
participation (Dwijuliani et al., 2021). Moreover, the current generation of learners, who are 
accustomed to digital devices and technological advancements, exhibited increased 
receptiveness to AI-based learning environments, corroborating previous studies 
emphasizing the importance of aligning educational approaches with learners' technological 
backgrounds (Widodo et al., 2021). 

Teachers acknowledged the practical benefits of utilizing Hume AI for assessing 
speaking skills. Compared to traditional, predominantly manual assessment methods—such 
as memorized dialogues, debates, or oral presentations—Hume AI offers real-time, 
interactive speaking tasks that better reflect authentic language use. This aligns with 
Brown’s (Brown, 2007) emphasis on assessing fluency, accuracy, pronunciation, grammar, 
vocabulary, and gesture through spontaneous speaking activities. Furthermore, teachers 
perceived that AI-based assessment could help prepare students for real-world applications, 
such as job interviews, consistent with the curricular goals of senior high schools. 

Nevertheless, several challenges emerged that may hinder the optimal integration of 
Hume AI in speaking assessment. The most significant barrier relates to students' 
insufficient English language proficiency, particularly in oral communication. This finding 
resonates with previous research emphasizing speaking as one of the most challenging 
language skills to master and assess (Navidinia et al., 2019; Zaim, 2017). Students' limited 
vocabulary, lack of listening comprehension, and unfamiliarity with spontaneous interaction 
in English created obstacles during AI-assisted assessments. Furthermore, the 
socioeconomic background of students, characterized by low-income families and limited 
access to private English courses, exacerbated these challenges, reducing students' 
motivation and preparedness for speaking tasks—a phenomenon previously observed in 
similar educational contexts (Jo, 2024). 

Additionally, this study highlights infrastructural constraints as a major limitation. 
Inadequate facilities, such as insufficient internet connectivity, lack of supporting devices 
(e.g., speakers, televisions), and outdated technological infrastructure, impede the effective 
use of AI-based assessment tools. This finding echoes Helmiatin's (Helmiatin et al., 2024) 
argument regarding the unequal distribution of technological resources in Indonesian 
schools, which restricts the widespread adoption of AI in education. 

In line with Kim (Kim, 2021), it is imperative for educational stakeholders to adopt a 
systematic, context-sensitive approach to AI integration. Such an approach should prioritize 
not only the technological dimensions but also the pedagogical, infrastructural, and socio-
economic factors that influence the successful application of AI in language assessment. 

Overall, while the integration of Hume AI in speaking assessment demonstrates 
promising potential to enhance student engagement, motivation, and authentic assessment 
practices, its effectiveness is contingent upon addressing several contextual barriers. These 
include improving students' English language foundation, increasing access to 
supplementary language learning opportunities, enhancing technological infrastructure, and 
providing adequate teacher training for AI-assisted assessment implementation. 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

176 
 

 
CONCLUSION 

This study revealed that the integration of Hume AI has a positive impact on enhancing 
speaking assessment practices by fostering student enthusiasm, increasing motivation, and 
providing a more authentic and interactive assessment experience. 
Teachers reported that students demonstrated increased interest and active engagement 
when Hume AI was introduced, highlighting the relevance of integrating technology into the 
learning and assessment process to align with students' familiarity and comfort with digital 
tools. Moreover, teachers acknowledged the potential of Hume AI to support the 
achievement of curricular goals, particularly in preparing students to use English fluently for 
real-world communication, including future employment opportunities. 

However, several challenges hinder the optimal use of Hume AI in speaking 
assessments. The most prominent challenges include students limited English proficiency, 
particularly in spontaneous speaking, their low motivation to learn English due to socio-
economic factors, and insufficient access to technological infrastructure within schools. 
These issues reflect broader systemic constraints that must be addressed to ensure the 
effective integration of AI in educational settings. 

Therefore, while Hume AI offers promising opportunities to enhance speaking 
assessment in senior high schools, its successful implementation requires comprehensive 
efforts to improve students' language competence, strengthen their motivation, and provide 
adequate technological support. 

ACKNOWLEDGMENT 

With profound respect and heartfelt gratitude, the author extends her deepest 
appreciation to her beloved parents for their unwavering love, steadfast support, sincere 
prayers, and constant encouragement, all of which have served as a strong foundation 
throughout the research process. The author also wishes to convey sincere thanks to the 
supervisor for her invaluable guidance, constructive feedback, and insightful contributions 
at every stage of this study. May all their kindness and support be rewarded abundantly by 
Allah SWT. 

REFERENCES 

Alsalem, M. S. (2024). EFL teachers’ perceptions of the use of an AI grading tool (CoGrader) 
in English writing assessment at Saudi universities: an Activity Theory Perspective. 
Cogent Education, 11(1). https://doi.org/10.1080/2331186X.2024.2430865 

Anggreini, S., Musiman, M., & Pratiwi, D. S. (2023). Teachers’ and Students’ Perception on the 
Use of Cake Application toward Their Speaking Skill On Merdeka Belajar Era. ELT-
Lectura, 10(2), 133–142. https://doi.org/10.31849/elt-lectura.v10i2.15159 

Anton, M. (2009). Dynamic Assessment of Advanced Second Language Learners. Foreign 
Language Annals, 42(3), 576–598. https://doi.org/https://doi.org/10.1111/j.1944-
9720.2009.01030.x 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

177 
 

Aprianoto, & Haerazi. (2019). Development and assessment of an interculture-based 
instrument model in the teaching of speaking skills. Universal Journal of Educational 
Research, 7(12), 2796–2805. https://doi.org/10.13189/ujer.2019.071230 

Beaumont, C., O’Doherty, M., & Shannon, L. (2011). Reconceptualising assessment feedback: 
A key to improving student learning? Studies in Higher Education, 36(6), 671–687. 
https://doi.org/10.1080/03075071003731135 

Berry, D. M., & Dieter, M. (2015). Thinking postdigital aesthetics: Art, computation and 
design. Postdigital Aesthetics: Art, Computation and Design, 1–11. 
https://doi.org/10.1057/9781137437204_1 

Brown, H. D. (2007). Principle of Language Learning and Teaching. In Encyclopedia of the 
Sciences of Learning (Fifth Edit). Pearson Longman. 

Budiharto, R. A., & Affandi, A. (2018). The Roles of Teacher in Teaching English Based on 
School Based Curriculum at the VII A Grade of SMPN I Kadur Pamekasan. Interaksi Jurnal 
Kependidikan, 13(1), 44–52. 

Crick, R. D., & Yu, G. (2008). Assessing learning dispositions: Is the Effective lifelong learning 
inventory valid and reliable as a measurement tool? Educational Research, 50(4), 387–
402. https://doi.org/10.1080/00131880802499886 

Dwijuliani, R., Rijanto, T., Munoto, Nurlaela, L., Basuki, I., & Maspiyah. (2021). Increasing 
student achievement motivation during online learning activities. Journal of Physics: 
Conference Series, 1810(1). https://doi.org/10.1088/1742-6596/1810/1/012072 

Fitria, T. N. (2025). Artificial Intelligence (Ai) in Education: Using Ai Tools for Teaching and 
Learning Process. IJIREEICE, 13(2). https://doi.org/10.17148/ijireeice.2025.13206 

Gorecky, D., Schmitt, M., Loskyll, M., & Zühlke, D. (2014). Human-machine-interaction in the 
industry 4.0 era. Proceedings - 2014 12th IEEE International Conference on Industrial 
Informatics, INDIN 2014, 289–294. https://doi.org/10.1109/INDIN.2014.6945523 

Hattie, J. (2012). Visible Learning for Teachers : Maximazing Impact on Learning. Routledge. 
https://doi.org/10.4324/9780203181522-6 

Helmiatin, Hidayat, A., & Kahar, M. R. (2024). Investigating the adoption of AI in higher 
education: a study of public universities in Indonesia. Cogent Education, 11(1). 
https://doi.org/10.1080/2331186X.2024.2380175 

Hung, S. T.-A., & Huang, H.-T. D. (2015). Blogs as a learning and assessment instrument for 
English-speaking performance. Interactive Learning Environments, 24(8), 1881–1894. 
https://doi.org/10.1080/10494820.2015.1057746 

Jo, H. (2024). From concerns to benefits: a comprehensive study of ChatGPT usage in 
education. International Journal of Educational Technology in Higher Education, 21(1). 
https://doi.org/10.1186/s41239-024-00471-4 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

178 
 

Kim, J. (2021). A New Era of Education: Incorporating Machine Teachers Into Education. 
Journal of Communication Pedagogy, 4, 121–122. 
https://doi.org/10.31446/jcp.2018.02 

Kustyarini, K., Utami, S., & Koesmijati, E. (2020). The Importance of Interactive Learning 
Media in a New Civilization Era. European Journal of Open Education and E-Learning 
Studies, 5(2), 48–60. https://doi.org/10.46827/ejoe.v5i2.3298 

Lasi, H., Fettke, P., Kemper, H. G., Feld, T., & Hoffmann, M. (2014). Industry 4.0. Business and 
Information Systems Engineering, 6(4), 239–242. https://doi.org/10.1007/s12599-
014-0334-4 

Mundy, K., Bickmore, K., Hayhoe, R., Madden, M., & Madjidi, K. (2008). Comparative and 
International Education : Issues for Teachers. Canadian Scholars. 

Mutia, L., Gimin, G., & Mahdum, M. (2020). Development of Blog-Based Audio Visual Learning 
Media to Improve Student Learning Interests in Money and Banking Topic. Journal of 
Educational Sciences, 4(2), 436. https://doi.org/10.31258/jes.4.2.p.436-448 

Naibaho, L. (2019). Teachers’ Roles on English Language Teaching: a Students Centered 
Learning Approach. International Journal of Research -GRANTHAALAYAH, 7(4), 206–
212. https://doi.org/10.29121/granthaalayah.v7.i4.2019.892 

Navidinia, H., Mobaraki, M., & Malekzadeh, F. (2019). Investigating the effect of noticing on 
EFL students’ speaking accuracy. International Journal of Instruction, 12(1), 83–98. 
https://doi.org/10.29333/iji.2019.1216a 

Patton, M. Q. (2002). Qualitative Research & Evaluation Methods (Thrid Edit). Sage 
Publications. 

Poudel, N. R. (2022). Different Roles of Teachers in the ELT Classroom. Journal of NELTA 
Karnali, 1(November), 17–23. https://doi.org/10.3126/jonk.v1i1.57224 

Rahmawati, Y., & Ertin. (2014). Developing Assessment for Speaking. Indonesian Journal of 
English Education, 1(2), 199–210. https://doi.org/10.15408/ijee.v1i2.1345 

Safdari, M., & Fathi, J. (2020). Investigating the role of dynamic assessment on speaking 
accuracy and fluency of pre-intermediate EFL learners. Cogent Education, 7(1). 
https://doi.org/10.1080/2331186X.2020.1818924 

Stenbom, S., & Geijer, L. (2025). Primary school teachers’ perception of digital 
transformation and their teaching role. Scandinavian Journal of Educational Research, 
69(5), 1131–1144. https://doi.org/10.1080/00313831.2024.2394395 

Tang, K. S., Cooper, G., Rappa, N., Cooper, M., Sims, C., & Nonis, K. (2024). A dialogic approach 
to transform teaching, learning & assessment with generative AI in secondary 
education: a proof of concept. Pedagogies : An International Journal, 19(3), 493–503. 
https://doi.org/10.1080/1554480X.2024.2379774 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689


  
 

Copyright © The Author(s) 
Vol.6, No. 3, July 2025 

e-ISSN: 2723-4126 
p-ISSN: 2776-8880 

 

179 
 

Vasagar, J. (2017). How robots are teaching Singapore’s kids. Financial Times, 1–7. 
https://www.ft.com/content/f3cbfada-668e-11e7-8526-7b38dcaef614 

Vercellotti, M. Lou. (2015). The development of complexity, accuracy, and fluency in second 
language performance: A longitudinal study. Applied Linguistics, 38(1), 90–111. 
https://doi.org/10.1093/applin/amv002 

Vial, G. (2021). Understanding digital transformation: a review and a research agenda. 
Routledge. 

Widodo, A. W., Solikhatun, I., Raharja, S., Salam, A. A., & Wartini, S. F. (2021). A Utilization of 
Information Technology on Education in Indonesia (2017-2020): A Systematic 
Literature Review. Journal of Physics: Conference Series, 1779(1). 
https://doi.org/10.1088/1742-6596/1779/1/012024 

Zaim, M. (2017). Implementing Scientific Approach to Teach English at Senior High School in 
Indonesia. Asian Social Science, 13(2), 33. https://doi.org/10.5539/ass.v13n2p33 

Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of 
research on artificial intelligence applications in higher education – where are the 
educators? International Journal of Educational Technology in Higher Education, 16(1). 
https://doi.org/10.1186/s41239-019-0171-0 

Zhu, M., Liu, O. L., & Lee, H. S. (2019). The effect of automated feedback on revision behavior 
and learning gains in formative assessment of scientific argument writing. Computers 
and Education. https://doi.org/10.1016/j.compedu.2019.103668 

 

http://u.lipi.go.id/1593190689
http://u.lipi.go.id/1593190689

