Lecturers perceptions of using Artificial Intelligence in Tertiary Education in Uzbekistan

Artificial intelligence (AI) has revolutionized different aspects of society, including higher education. This paper investigates faculty members' perceptions of using artificial intelligence, chatbots, and generative AI in teaching and learning in Uzbeki's higher education contexts. The researchers employed a qualitative study using semi-structured interviews to collect the data. Purposeful sampling was employed to select participants in this study, and the interviews were conducted face-to-face at the university campus. The study's underlying theory is the Unified Theory of Acceptance and Use of Technology (UTAUT) model, including effort expectancy, performance expectancy, social influence, and facilitating conditions, which were employed as a lens to direct the research. The data were transcribed and analyzed using the deductive approach for the thematic analysis. The findings revealed that lecturers have a positive attitude to adopt and use AI for content creation, assessment and feedback, and doing research in their institution. Some instructors may see it as a valuable tool for generating creative content and aiding student learning. In contrast, others may have concerns about its potential to replace human creativity or biases in generated materials. Lecturers also view AI as a technology to achieve accessibility and equity after overcoming the challenges. Findings revealed that some measurements should be taken about the facilitating conditions and the perceived risks of using AI.


INTRODUCTION
In recent years, artificial intelligence (AI) has revolutionized various aspects of society, including higher education.This transformative technology is actively reshaping the educational landscape by automating tasks, tailoring learning experiences, and extracting valuable insights from student data.As AI continues to become more prominent in higher education, it becomes crucial to examine its perception among stakeholders.Consequently, this paper aims to investigate how lecturers in Central Asia perceive artificial intelligence within the higher education context, with a specific focus on Uzbekistan.To achieve this objective, the paper will address the following research question: The subsequent section will offer an overview of the background and existing literature on the subject.The paper will then explain its methodological approach and present the outcomes of conducted interviews.A nuanced discussion of the findings will be provided in context, accompanied by an exploration of the theoretical and practical implications related to the integration of AI in education.Finally, the paper will conclude by acknowledging the limitations of this research and outlining avenues for future work.

LITERATURE REVIEW
To grasp the perception of AI among stakeholders in Central Asia, it's crucial to establish a shared understanding by examining the existing key themes in the integration of Artificial Intelligence (AI) in education and the current perspectives of AI worldwide.

KEY THEMES IN INTEGRATING ARTIFICIAL INTELLIGENCE (AI) IN EDUCATION 3.1 Adaptive and Personalized Learning
Adaptive learning, also known as intelligent tutoring systems (ITS), is a virtual learning environment that tailors teaching and learning methods to individual learners' skills and needs.It incorporates machine learning techniques, self-training algorithms, and neural networks to determine the appropriate learning content for students (JISC, 2021).This personalized approach helps identify each student's proficiency level and provides relevant activities and assessments.Adaptive learning systems, AI-assisted marking and feedback, chatbots, and virtual teaching assistants are examples of how AI is used in education.For instance, Arizona State University implemented CogBooks, an adaptive learning system that replaced traditional textbooks and resulted in improved passing rates and reduced dropout rates (JISC, 2021).Other examples include TSAL and WELSA, which integrate various AI tools like adaptive learning, automated feedback, grading, and chatbot responses.Another intelligent tutoring system, Jill Watson, combines AI technologies with Pearson Education specialists to create a virtual teaching assistant that engages students through tailored conversations and personalized learning.These AI-powered systems offer a gateway to accessible higher education beyond physical institutions.(JISC, 2021).

Automated Grading and Feedback
The use of artificial intelligence (AI) in grading students' work is becoming increasingly common.One approach is the autograder program, which can evaluate written assignments and multiplechoice tests without human involvement (August & Tsaima, 2021).Another example is a system that combines AI and virtual reality to assess dentistry students' proficiency based on their movements.AI-powered tools like Gradescope are also being used to digitize and assess students' assignments, saving instructors time and effort.However, some argue that relying too heavily on AI grading can stifle students' creativity and lead to uniformity.Despite these concerns, automated essay-scoring AI systems have been widely used in large-scale assessments like the SATs, offering time and cost saving (Braiki et al., 2020).Platforms like edX have developed AI scoring engines that aim to mimic human grading, providing feedback on a large volume of writing assignments.While this approach may be effective in online learning environments, it may restrict students' ability to express their own ideas in traditional classrooms (Baker, 2021).

Emotional AI
Education Technology companies are using Emotional AI to measure social and emotional learning (McStay, 2019).Affective computing is a field of study that focuses on building systems and devices that can detect and interpret human emotions.Autonomous engagement monitoring systems track and report student engagement levels using nonverbal signs analyzed by AI technologies (Pabba & Kumar, 2021).Studies have shown that AI can accurately detect student engagement through eye and head movements, facial emotions, and affective states.This tool can be used to enhance engagement and achieve learning results, but there are privacy concerns and limitations for students with certain disabilities (Pabba & Kumar, 2021).

AI for Equity and Accessibility
AI systems utilizing machine learning algorithms have the potential to provide equal learning opportunities for all students, regardless of their abilities or backgrounds (Holstein & Doroudi, 2021).Governments are encouraged to prioritize equity and inclusion in AI education policies, particularly for marginalized communities.While initiatives like IBM's Simpler Voice: Overcoming Illiteracy Project aid in accessibility, challenges such as fairness, accountability, and transparency in AI systems remain.Factors such as limited access to technology and language proficiency can also create disadvantages for certain groups (UNESCO, 2019).The use of AI in education has the potential to increase rather than reduce disparities in education for marginalized groups, as well as reinforce biases and inequalities.AI-based solutions can offer personalized learning experiences and address barriers faced by disadvantaged groups.However, if the algorithms are biased, the outcomes will also be biased.Microsoft's Tay, a Twitter chatbot, is an example of this (Dhawan and Batra, 2021).Therefore, ethical considerations and stakeholder engagement are necessary to ensure the responsible use of AI in education.Policymakers must prioritize equity, inclusion, and accessibility to maximize AI's potential in promoting equal and inclusive education (UNESCO, 2021).

International Perspectives on AI in Education
Based on the current research, it is evident that perceptions towards integrating AI into educational practices vary among educators globally.For instance, according to a study conducted in Estonia, K-12 teachers expressed that they have limited AI knowledge but have positive attitudes toward its potential.Moreover, they believe that AI is a valuable support tool for tasks like retrieving learning materials and organizing lessons.(Chounta, 2022).In a separate Estonian study, university-level teaching staff were optimistic regarding emerging technologies (ETs).The participants of the study also emphasized the importance of prior ET-related experience in fostering positive attitudes and successful implementations in higher education (Leoste, 2021).In Asia, educators in Pakistan also have positive perceptions of technology integration in teaching as they believe that the tool will help them enhance instructional practices and increase student motivation (Akram et al., 2022).Meanwhile, in South Korean prospective, teachers recognize the demand for AI in school mathematics but are concerned about its potential to undermine students' independent thinking (Shin., 2020).Other concerns have been seen in studies that were undertaken in Europe and North America.For instance, based on a quantitative study conducted in Sweden several university teachers reveal concerns about fairness, responsibility, and insufficient knowledge and resources for engaging with AI in teaching practices (McGrath et al., 2023).When it comes to North America, concerns arise regarding the role of teachers and the transparency of AI decision-making are raised by STEM educators (Kim, 2022).
The integration of AI into education requires context-specific considerations as it is important to acknowledge and address the challenges unique to each region to ensure effective and equitable deployment of AI in education worldwide.Thus, collaborative efforts informed by regional insights and global best practices will play a pivotal role in shaping the future of education.

PROBLEM STATEMENT
Due to the technological advancements in large language models (LLMs), new generations of chatbots are highly competent in understanding users' intentions, empowering natural speech conversation, and providing personalized, interactive, and affordable support for the user (Akinwalere & Ivanov, 2022; Deng & Yu, 2023).In tertiary contexts, these chatbots provide personalized feedback, answer students' questions, and establish communication among instructors and students.However, more is needed to know about AI integration in Uzbekistan's higher education.It is important to investigate how AI is used in Uzbeki higher education for several reasons.Uzbekistan has limited resources in terms of databases and access to quality education.Integrating AI, particularly educational AI, such as virtual teacher aids, smart campus planning, personalization, and automation, can be an opportunity to overcome shortcomings.Another reason is that there is no literature regarding the cultural adaptation of AI in higher education and how far AI apps fit into the cultural circumstances of Uzbekistan (UNESCO, 2023).Moreover, it is vital to evaluate the readiness and training of Uzbek university lecturers on AI inclusion in education.It is essential to determine if faculties consider themselves appropriately prepared to use AI products in teachers' practice successfully.Another gap is to analyze the ethical concerns of applying automated grading schemes in Uzbekistan.There is a need to investigate the lecturer's concerns on fairness, bias, and transparency in AI-based grading.

Research Questions
1-How do instructors perceive the benefits and drawbacks of employing generative AI in higher education in Uzbekistan?
2-What are the instructors' perceptions of using generative AI and chatbots in higher education in Uzbekistan, specifically in teaching and learning, lesson planning, personalized learning, giving feedback, automated grading, and emotional AI?

Conceptual Framework
The unified Theory of Acceptance and Use of Technology (UTAU) proposed by Venkatesh et al. (2003), illustrates factors influencing users' adoption of technology and was chosen for this study.This theory was originally introduced to explain user behavior and intentions in information system contexts.It is worth mentioning that this model has been broadly validated and applied across different technological contexts (Slepankova, 2021).
The UTAUT model comprises several key constructs that affect users' acceptance and use of technology.Performance expectancy refers to the extent to which users think that technology will assist them in performing tasks more effectively.Effort expectancy refers to the perceived ease linked with technology use.
Social influence discusses the effect of social factors, subjective norms, and the effect of others on users' technology acceptance.Facilitating Conditions is defined as the extent to which users judge that the technical and organizational infrastructure supports the use of technology.These four main constructs are determinants for the users to adopt the technology.
The UTAUT model has four other mediating variables, including gender, age, experience, and voluntarism of use, that may influence the four main constructs (Venkatesh et al., 2003).For instance, age plays a major role in adopting a technology.Furthermore, the amount of experience can influence technology adoption as an individual with great experience could be ready to adopt new technologies.In contrast, individuals with limited experience may find it more challenging to adopt the technology.UTAUT model has been employed with various technologies such as cell phones, e-commerce, social media, and medical equipment (Venkatesh et al., 2003).
The current study extends the UTAUT model in adopting AI in higher education contexts to examine the aspects influencing AI acceptance and use among university lecturers.Regarding performance expectancy in AI integration in higher education, this study investigates lecturers' perceptions of AI's ability to facilitate administrative tasks, teaching, and learning, which will significantly influence their adoption of AI, referring to enhanced student engagement, efficient grading and feedback, and personalized learning experiences.Effort expectancy in AI adoption refers to instructors' perceptions of how easy it is to integrate AI into their current practices.This construct involves the compatibility, simplicity, and user-friendliness of AI applications.
Social influence refers to the influence of administrators, peers, and institutional culture on instructors' decisions to adopt AI in academic settings.Social factors include perceived expectations and support from superiors and colleagues in higher educational contexts.
Facilitating conditions discusses lecturers' perceptions of the technical and organizational support for integrating AI into higher education, which include access to professional development, training, and resources that will affect their readiness to integrate AI technologies into their teaching.
Perceived risk (Bauer, 1960) was taken for this study's purpose and included in the model as a fifth category.Perceived risk was described by Bauer (1960) as the degree of uncertainty and possible drawbacks that a buyer attributes to a specific buying choice.Although the term is primarily used in marketing, as it is a significant factor that can influence the acceptance of AI, it is being applied in this study to refer to the risks and ethical issues related to using AI in higher education.Since integrating AI in higher education is a significant challenge, this connects to all research questions.
Perceived risk has been included as a new construct to the model, which is still applicable to adopting technology and appropriate for this study.Equity and accessibility were included under the social influence category, but they may have independently established a new domain.Thus, the model might receive suggestions in the future.The theoretical framework of the study has been illustrated in Figure 1 below: The data were coded and grouped into themes using the constructs established in the newly suggested framework.Furthermore, the findings were guided by the constructs in the discussion section.

Method
Due to the complexity of the AI phenomenon in the context of higher education, this study adopted a qualitative design to provide more significant and in-depth insights into integrating AI in higher education and the faculty members' experiences and viewpoints on using AI.As Plano and Creswell (2015) pointed out, qualitative research examines a complex phenomenon by exploring an individual's point of view and the essence of lived experience (Creswell & Poth, 2017).In the current study, the concept of experience is based on John Dewey's description, an experience that refers to developing organism-environment relationships (Acampado, 2019).
People's experiences constantly change or become dynamic as they interact with their environment (Acampado, 2019).Therefore, one's experience differs from another as it considerably depends on one's new interaction with their environment.In this study, the researchers examine the faculty members' lived experience and how they interact with their environment and AI applications and investigate their standpoints of AI in their environment.The semi-structured interview was chosen because this approach allows the researcher to understand faculty members' experience in integrating AI into higher education.Secondly, most research questions are exploratory so that participants' responses can lead to future research and assist the researcher in developing a more thorough knowledge for future research.

DATA COLLECTION
The first step in this process is to develop interview questions based on the topics reviewed in the literature to guide the interview while allowing flexibility for examining and follow-up questions.The interview guide was validated by two experts in the field to ensure the questions were appropriate, reliable, and consistent for this study.The experts provided valuable feedback on the clarity of the questions to enhance the relevance and validity of the research.It has been verified that the interview protocol questions aligned with the respondents' experience.
Participants of this study were fifteen faculty members teaching at different faculties, such as engineering, business, medical, dentistry, and English education.Purposeful or purposive sampling was employed as the sampling method.The researchers explained the purpose of the research to participants, and they obtained consent from each participant.
The researchers scheduled and conducted one-on-one semistructured interviews.Following each open-ended question, participants were asked to respond to follow-up questions regarding integrating AI into higher education and their perceptions of using AI.This method allowed the researchers to gain in-depth and accurate information about the topic (Plano & Creswell, 2015).All interview sessions were recorded and transcribed for further analysis.

Setting
The interviews were conducted face to face on Central Asian University's campus site.This approach allows the researchers to read the interviewees' body language and encourage them to interact with them actively.The interviews continued from half an hour to an hour.

Participants
In the current study, purposeful or purposive sampling was employed as the sampling method.Purposeful sampling is a nonprobability method in which participants are selected based on specific characteristics related to the study's objectives (Plano & Creswell, 2015).The researchers selected lecturers from different faculties at Central Asian University.Three lecturers were selected from the faculty of engineering, three from the Department of English Education, two from the faculty of dentistry, two from the faculty of medicine, and five lecturers were selected from the school of business, comprising 15 participants.The lecturer's information is illustrated in Table 1.For anonymity and confidentiality, the researchers chose pseudonyms for the participants.

Data Analysis
After transcribing the audio recording into the text using manual verbatim or word-for-word transcription, the researchers scanned the transcript to familiarize themselves with the interview content.

Challenges in AI implementation
In the next stage, the data were annotated and open-coded initially.The researchers then started to do axial coding to refine the emerging categories.(Korstjens, & Moser, 2018).The deductively coded data were organized through axial coding to identify the relationships and connections between the predefined categories.At this stage, the researchers found some categories that needed to be specified under the new subcategories.Following this, overarching ideas related to the integration of AI were identified, and related categories were grouped under the themes that summarize the broader concepts that emerged from deductive coding.In the current study, deductive coding was employed as the codes were deduced from the theoretical framework, which refers to performance expectancy, social influence, effort expectancy, facilitating conditions, and perceived risks.
Consequently, new data with predefined codes and themes were constantly compared to ensure the deductive approach aligned with the nuances of the data.At this point, member checking was conducted by asking the participants to validate the deductively derived themes to ensure the alignment between their responses and predefined themes.In the final stage, findings were reported comprehensively using deductively derived themes and any emergent themes that arose during analysis.Table 2 illustrates a sample of coding scheme.

FINDINGS
This section presents research findings of faculty members' perspectives in higher education institutions about using artificial intelligence for teaching and learning, as well as how it can contribute towards equity and accessibility.The challenges posed by using AI in higher education in Uzbekistan are also highlighted in this section.The challenges posed by using AI in higher education in Uzbekistan are also highlighted in this section.The findings are presented in conjunction with the themes based on the theoretical framework: performance expectancy, effort expectancy, social influence, facilitating conditions, and perceived risk.
Of 15 faculty members, 13 lecturers were active users of AI.It demonstrates that the majority of instructors were familiar with the applications of AI at different schools.
The role of artificial intelligence in teaching and learning at ACU will be highlighted in this section.

Proofreading
The majority of lecturers mentioned that they frequently use Chat GPT and grammar checkers to proofread their research.
Lecturer 8 stated that: "The integration of tools like chat GPT and grammar checkers, such as CoolSpot, aids in expediting processes like research paper writing and grammar verification."

Doing research
Figure 2: The role of AI in teaching and learning Another lecturer pointed out that he found Chat GPT helpful in doing research.
"I frequently use AI for analyzing the data, but bear in mind that we need to give AI some accurate prompts to provide us with precise coding and themes in qualitative analysis".Assistance with coding and advanced statistical support Lecturer 14 in the School of Computer Engineering highlighted the role of AI in assisting students with coding and statistical analysis.
"AI tools like Chat GPT offer customized programming codes upon request, assisting students in data science tasks and statistical analyses.""In the data science realm, AI tools contribute to advanced statistical analysis, expanding students' capabilities beyond traditional methods." Problem-solving tool Another professor maintained that AI is implemented by students for preparing exams and as a problem-solving tool.
"Students leverage AI for exam preparation, seeking assistance in solving problems and gaining a deeper understanding of course materials.

Performance Expectancy
The following section presents findings on AI performance expectancy, which refers to the extent to which the use of AI may result in enhanced job performance by the faculty member and better student learning opportunities.
Chatbots and Digital Assistants Most faculty members (lecturers 4,9,10,13,12) stated that one of the main positive aspects of chatbots and digital assistants is that they are widely accessible and provide rapid guidance and response.Lecturer 6 elaborated that lecturers cannot permanently answer students' inquiries.Lecturer 8 also mentioned that she might be in a meeting and unable to respond to students' questions, so students may consult with chatbots to facilitate their learning.
Similarly, lecturer 10 stated that chatbots could help students perform low-level tasks that require little skill, whereas faculty members can conduct high-order tasks.He referred to Bloom's taxonomy, highlighting that the chatbot could help students reach the lowest taxonomy level.
Lecturer 4 elaborated on how digital assistants and chatbots can provide prompt feedback on assignments, quizzes, and practice exercises.This immediate feedback improves the learning process.
Moreover, lecturer 9 maintained that: "Chatbots are everywhere.They can assist us in doing routine administrative tasks in higher education, such as course registration, enrollment, employability, giving information about the academic calendar, and so on.He believes that this restructures administrative processes and frees up time for instructors." Lecturer 13 further supported this point by asserting that "Chatbots can serve as friendly companions helping learners in decision-making as an academic advisor.Faculty 10 mentioned that they increase students' engagement in a human-like way.Once the chatbots have solved their academic problems, consequently they will be more engaged in academia".
However, six faculty members, Lecturers 1, 3, 5, 7, 8,14, emphasized human interaction in addition to consulting with chatbots and digital assistants.Lecturer 7 stated that: "Interaction and connection among lecturers and faculty members are essential, and we cannot only count on these Chatbots." Faculty 3 also highlighted this point by mentioning that these technologies prevent students from thinking critically and creatively.
"A constant consultation with chatbots and digital assistants makes us lazy, so we are reluctant to think out of the box.These technologies inhibit creativity." Lecturer 1 highlighted the complexity of the educational context and the inability of chatbots to understand the context and nuances of conversation, which ends up frustrating the user.
Lecturer 7 highlighted that another disadvantage of chatbots is their lack of emotions.AI-driven chatbots frequently lack empathy, which can be a significant drawback in educational settings.They might struggle to recognize and respond appropriately to users' emotional needs or states.
Adaptive and Personalized Learning All faculty members agreed that tailoring adaptive learning to suit students' pace and level is advantageous.Some lecturers highlight advantages, such as providing immediate feedback, customized content, enhanced engagement, efficient time management, identifying learning gaps, and considering students' learning styles.
On the other hand, some lecturers highlighted that AI-driven personalized learning comes with potential disadvantages such as data privacy and misuse of sensitive information, overemphasis on assessment, lack of holistic assessment, lack of human connection, and technological barriers, costs, and implementation costs.Lecturer 7 maintained that AI-enabled personalized learning is unsuitable for all courses and will work better only for online selfpaced courses.
Automated Feedback and Automated Grading Almost all faculty members use automated grading, and they believe that this way of grading is time-saving and energy-saving However, in terms of automated feedback, lecturer 4 stated that: "As a human being, you can see the bigger picture of what the students want to convey, so you can at least give them partial credit.However, an AI will not be able to make this judgment." Some lecturers maintained that: "Human interaction in providing feedback is vital, but students must understand why they got this grade; AI cannot explain the decision they made."

Effort Expectancy
This section includes the faculty member's efforts to adopt AI.This is closely related to determining whether they are willing or not to accept AI according to their effort and experiences.
A significant challenge almost all faculty members mentioned is the need for more professional development training for AI adoption at the university.Some lecturers 8,6, 2 asserted that the adoption of AI largely depends on their background and faculty.Older faculty members need to become more familiar with the existing chatbots and educational assistance; they value the traditional methods; therefore, it is evident that they need more effort to be able to adopt AI.
Social Influence Social influence refers to the degree to which others use AI and its popularity.Achieving equity and accessibility is discussed in this section.AI is becoming more popular and is being used by many people across the globe.Most lecturers believe that AI is here to stay and that they must embrace it.Higher education is no exception; they consider it an opportunity, not a threat.
Lecturer 2 stated that: "It is a phenomenon that cannot be avoided anymore, and we should embrace AI wholeheartedly due to its potential ".
Lecturer 10 also stated that: "AI adoption is essential for students to have skills that match the job market." Regarding equity and accessibility, most faculty members were hesitant as accessing Chat GPT is not available in Uzbekistan, while it is accessible in other countries.However, they use other open AI that is freely accessible in Uzbekistan.

Facilitating Conditions
This section presents the findings related to facilitating conditions, including infrastructure, resources, and policy documents, for adopting AI in higher education.All lecturers maintained that their university has no policy document, particularly for AI.They clarified that the university announced an academic integrity policy; however, nothing was detailed for AI.
Most faculty members maintained that the university must arrange some workshops to raise faculty members' awareness of how to detect AI-generated texts.
Perceived risk Perceived risk refers to the degree to which one perceives that the use of technology might lead to negative results or harm.All lecturers approved that the main challenges of using AI are bias, data protection, surveillance, and privacy.Faculty 10 demonstrated that: "Everything associated with AI requires data, so confidentiality and privacy are the main concerns while using AI with students in education".
However, lecturer 15 mentioned that: "As a professor specializing in AI, I do not like to look at the cons because I believe those cons can be eliminated using many tools and applications".
Lecturer 15 also stated that "using emotion AI surveillance and being watched makes the students behave differently and will change how they behave and project their emotions in the classroom".
Despite discrimination, privacy, surveillance, confidentiality, bias, and data protection issues, some faculty members were still optimistic about AI adoption and proposed some solutions to overcome these challenges.

DISCUSSION AND CONCLUSION
This study set out to investigate lecturers' perceptions of AI and the adoption of AI in higher education in Uzbekistan.The results of this study showed that IA support and adoption can assist in accomplishing the Sustainable Development Goals (SDGs), including infrastructure development, quality education, innovation, and poverty reduction.Faculty members explained that AI adoption is necessary in all sectors and can enhance students' skills, engagement, and employability.Faculty members maintained that AI chatbots and assistants are accessible to assist students in acquiring low-order skills so faculty can help them develop high-order skills.Furthermore, the findings were utilized to contextualize AI use in higher education in Uzbekistan, and subsequently, the lecturers are ready to embrace AI with various degrees.
Research Question 1 How do instructors perceive the benefits and drawbacks of employing generative AI in higher education in Uzbekistan?
University instructors' perceptions of generative artificial intelligence in higher education can vary.Some instructors may see it as a valuable tool for generating creative content and aiding student learning.In contrast, others may have concerns about its potential to replace human creativity or biases in generated materials.Understanding these perceptions is crucial for successfully integrating and accepting AI in higher education settings.
Most lecturers have viewed AI as an exciting opportunity for themselves and students to develop their creativity, explore new ideas, and collaborate with their peers more effectively.However, some lecturers were concerned about students' overreliance on AI or using it to undermine their critical thinking, creativity, and the learning process.They are also concerned about ethical issues such as plagiarism or data misuse by automated systems.These results are supported by the current literature (Popenici and Kerr, 2017).
Research Question 2 What are the instructors' perceptions of using generative AI and chatbots in higher education in Uzbekistan, specifically in teaching and learning, personalized learning, giving feedback, automated grading, and emotional AI?
The results of an interview conducted among instructors in higher education on their perceptions of using generative AI and chat GPT technology showed mixed reactions.Some instructors saw positive potential for using these technologies, such as allowing them to cover more material in less time and improve student engagement.Others were concerned with the quality of the content produced by these technologies, citing concerns about accuracy or relevance.Additionally, some instructors expressed concern that this type of technology could be used to replace faculty members' roles, creating job insecurity or reducing instructor autonomy.Finally, there was also some discussion around privacy issues associated with AI technology and the potential need for additional oversight regarding student data collection.The results of this study are supported by (Ivakhnenko & Nikolskiy, 2023; Walczak & Cellary, 2023).
The use of generative AI and chat GPT in higher education has been met with various perceptions from instructors.Some have embraced the new technologies, seeing them as valuable tools that can be used to support student learning and performance.For example, generative AI can provide personalized feedback on assignments or help students generate ideas for projects.Chat GPT technology can facilitate online discussions between instructor and student, providing immediate responses to questions and allowing conversations to flow more naturally.
On the other hand, concerns have also been raised about how this kind of technology may affect student-instructor relationships.Some worry that using automated systems could result in a lack of human connection between teacher and pupil, reducing opportunities for meaningful dialogue or shared understanding, which is integral to effective teaching practices.

Implication of the study
As per the UTAUT model (which Venkatesh and others suggested in 2003) and our research findings, college professors display high degrees of performance expectation, anticipated effort, and the effect of social factors.However, special points, mainly those focusing on enabling environments and apprehended risks, require thoughtful consideration.As a result, this research has various impacts on the proper deployment of AI in tertiary education.
The studies show that lecturers enjoy using AI in tertiary education despite some hurdles.This suggests a call for government plans to close the digital gap and guarantee fair internet and device access.Policymakers are advised to carve out an extensive plan fitting to the higher education situation in Uzbekistan.This should deal with the socio-economic issues and include points about inclusion, possible risks, and moral questions linked to AI in education.
Furthermore, the study suggests that AI has the potential to address challenges related to access to higher education.With the anticipated increase in demand surpassing the available supply, decision-makers can explore the development of AI-led learning opportunities for students.This proactive approach aligns with the participants' acknowledgment of AI as a tangible reality that institutions can either leverage or allow to pass by.

Recommendation for further study
Future studies may consider students' perspectives while gathering lecturers' opinions concerning AI use in teaching.Further research could investigate how learners view the incorporation of AI technology in their education, such as personalized learning, auto-grading, and feedback.This could be compared to the perspectives of the lecturers to establish any existing gaps.
Future studies could identify possible impediments to incorporating AI into Uzbekistan's higher education.These may entail technological infrastructure obstacles, policy hindrances, and barriers.Explore significant challenges that lecturers envision and how to navigate transitions for a smoother implementation of AI as we plan for the future.

Figure 1 :
Figure 1: Conceptual Framework of the study

Table 2 :
A sample of the coding scheme