Three layers of professional isolation – a graduate teaching assistant’s experience of grappling with AI use in college classrooms

Case Study | MediaEd Insights | AI: Educator Perspectives Edition | June 2026

Written by Salome Apkhazishvili

Salome Apkhazishvili is a Doctoral Candidate in the Department of Communication | Georgia State University

It is common knowledge that humans crave connection, and when they obtain it, they report better subjective health (Cacioppo & Patrick, 2009). As trivial as it may sound, those who compare conversational AI chatbots to human beings often call the authenticity of this evolutionary perspective on human connection into question. Although this is a pivotal topic of public conversations today, I am not equipped to offer a substantive take on it. In this case study, I would like to reflect on my experience of professional isolation as a graduate teaching assistant, grappling with my students’ use of AI tools during my final year of doctoral study in communication. I will first explain what professional isolation means in a higher education context and then illustrate it through my experience as a graduate teaching assistant, communications professional, and educator in its broad sense. This purely subjective reflection will be followed by questions inviting other graduate teaching assistants, educators, and communications professionals to share their thoughts as we navigate AI-driven educational and social landscapes.

Professional Isolation - Definition

Isolation is the opposite of connection, and it can occur at the physical, social, and psychological levels (Walters et al., 2024). When isolation happens in the work environment, individuals describe it as a lack of communication with colleagues and supervisors (Chekwa, 2018), a lack of opportunities for professional growth, and a lack of adequate resources to navigate challenging environments (Chekwa, 2018). According to Dussault and Thibodeau (1997), professional isolation refers to “the unpleasant experience that occurs when a person’s network of social relations at work is deficient in some important way, either quantitatively or qualitatively” (p.523). Flinders (1988) emphasizes the importance of differentiating between isolation as an actual condition and isolation as a perception. Studies on professional isolation in higher education have examined the experiences of adjunct faculty members (Buch et al., 2017), distance educators (Kanuka et al., 2008), preschool teachers (Aizenberg & Oplatka, 2019), principals (Dussault & Thibodeau, 1997), and individuals who experience an identity crisis in the teaching profession. A fear of losing control over teaching processes and the traditional educator role appears in recent empirical studies (e.g., Okulicz-Kozaryn et al., 2025) on educators’ perspectives on using AI as a teaching companion. Lack of adequate administrative and technical support for teachers is another concern highlighted in recent literature (Almuhanna, 2024). Although there are indications of professional isolation among teachers grappling with AI tools, the room for examining their experiences from a professional-isolation perspective remains large. Conducting studies on this topic is particularly important given research evidence of the direct negative effects of isolation on overall health and productivity among teaching professionals. In the meantime, I am initiating a conversation about it by sharing my personal experience grappling with the use of AI among my students as a graduate teaching assistant, communications professional, and educator. Hopefully, further discussion follows.

Layer 1 – Graduate Teaching Assistant
Being a graduate teaching assistant is a tricky role that varies across schools, departments, and professors; however, it is unsurprising that graduate teaching assistants have limited freedom. They may have freedom in how they interact with students, how they compose their feedback, or how they design lecture slides, but they cannot change assignments, especially when teaching basic courses that are often facilitated by multiple teaching assistants across multiple sections. For example, they may be allowed to slightly amend the assignment policy, with the reminder that this action may lead to confusion and frustration in other sections taught by other teaching assistants or by the course directors themselves. Similarly, they may have their own AI policy in a course syllabus, but it is strongly preferred to align with the university guidelines to prevent further chaos.

At the start of the fall 2025 semester, I received a link to the university-approved AI use guidelines (In 2025, these guidelines looked more abstract), which did not fully align with my teaching philosophy. While I was willing to consider the general guidelines in my course syllabuses, I knew that having conversations with senior faculty, administrators, and peers was necessary to seek more sensible approaches to AI use in the classroom. Motivated to dig deeper in search of better strategies, I reached out to my department, started a discussion prompt with fellow doctoral students who were assigned to teach the same courses, emailed the course directors, and requested a one-on-one consultation with teaching and learning department specialists. They responded. The grand themes from these conversations were as follows: “Our AI-screening tools are not sufficient enough”, “We should do our best to adhere to universal policies to avoid further chaos”, “Think of redesigning your assignments in a way that does not lead students to use AI tools”, “We cannot do much to prevent students from using AI.” I was processing these messages while grading speech outlines, case study proposals, reflection papers, and discussion posts partially or fully written by AI. My ad hoc solution was to ask students suspected of using AI to meet with me if they needed extra guidance on the assignment before resubmitting their original work. Almost none of these invitations were reciprocated. Some of them resubmitted their work, yet there was no significant improvement in originality. Some of them ignored the request and failed the class or received lower grades due to their overall performance over the semester. These were the signs that all key stakeholders in higher education – administrators, mentors, course directors, teaching assistants, and students – were not aligned and just following the inertia to close another challenging chapter in their academic year. While it is completely understandable that responding to the rapid integration of AI tools into the learning process is a complex task and that seeking solutions is time-consuming, I strongly believe that more collaborative approaches and peer- or group-support initiatives would make it more manageable for educators.

In addition to the lack of high-quality collaboration among those various players, I felt that my role as an instructor was fading. Very soon after starting my doctoral program in 2022, teaching became the shelter amid the perfect storm of cognitive and mental work required to complete coursework, conduct research, and handle grading responsibilities. I noticed that serving students as best as possible was becoming my primary identity in the program. I was willing to compose tailored feedback, meet with students as many times as they wanted for additional instruction, and try to understand their personal struggles while considering make-up work. Three years later, I first felt fear of losing control over this fulfilling process of being in conversation with students, hearing their needs, and figuring out responses to their challenges together. Suddenly, they stopped sending emails, asking to meet during office hours, and even complaining about my feedback. Reading AI-generated essays and reflections did not seem the worst part of it. What I was missing was that one-on-one mentoring where I thought I was doing something meaningful.

Being in a doctoral program is uncertain enough, but when you start to see the cornerstones of your strength and confidence crumble, the uncertainty widens further, leading to my next point about a professional identity crisis.

Layer 2 – Communications specialist

An article from Business Insider lists media jobs, such as advertising, content creation, journalism, and technical writing, among the professions at highest risk of being replaced by AI (Zinkula and Mok, 2024). While this popular narrative is yet to be supported by more rigorous predictive studies and qualitative inquiries, it surely casts doubt in the minds of junior scholars and professionals like me about the value of my field, which has been questioned in the political arena as well. So, even if I tried to make a case with my students about how valuable their voice is, I have an implicit fear of losing a battle with this story that AI is there to write emails, memos, and stories, so why spend the limited time and attention on using human voice doing the same tasks? Even after spending four years reading and learning about the major breakthroughs and contributions in the media and communication field, I don’t think I have a strong, persuasive counternarrative, and I would greatly appreciate it if educators spent more time developing such narratives and embracing the companion affordances of generative AI. Put differently, I would appreciate more diversity in AI-related discussions than I currently see. This sentiment is addressed in my next and final section.

Layer 3 – Educator

In each conversation I joined on LinkedIn or at informal gatherings like Media Education Lab virtual events, I noticed I was aligning myself with the techno-skeptic camp, even though I have always considered myself someone who embraces it. After all, my master’s thesis was about the collaborative use of smartphones to promote a better culture of parental mediation. After completing my MA in Communication Studies, I became a large-scale media literacy program manager in my home country, working with over 50 participants aged 16 to 30 and helping them see the power of media technologies and critically reflect on it. At the same time, I co-facilitated the Summer Institute in Digital Literacy, a professional development program for media literacy educators, professionals, and advocates. I also remained an active member of the Media Education Lab, which consistently gave me the opportunity to be inspired by other educators to use interactive new technologies in my own teaching. Later in my doctoral program, my research explored the digital divide and the optimization of digital tools for the older population. Yet, as I engaged with AI, I felt like I was dealing with a different type of media, but instead of figuring out why I was feeling that way, I immediately felt ashamed of being skeptical of AI use in the classroom. It is still mind-blowing to me that, in just three years of my doctoral program, we went from zero to widespread use of AI in the classroom. Building and implementing other technological innovations, such as the telephone, television, the internet, the personal computer, and the smartphone, took longer than the rates we see with AI adoption. Should not this stark difference in adoption rate be a telling sign that we may need more time to think before embracing it?

I am a serial overthinker, but I always know how to seek help, so here is an invitation to help me find a way back to students and educators who truly care for the profession and fully embrace the responsibility of guiding the next generation of humans.

Questions
Some of the questions below are better suited to group discussions, while others may easily turn into research inquiries. I look forward to hearing from you at salomeapkhazi@gmail.com or during our next Media Education Lab meeting, whatever shape it may take.

How do educators deal with professional isolation in the age of AI? What does that isolation look like? Is it a condition or a perception?
What are the effective ways of supporting graduate teaching assistants in the age of problematic AI use in classrooms? How do administrators and mentors ensure that the next generation of communication instructors is not overburdened and still finds their teaching role meaningful?
In the age of AI, where does the disconnect among higher education stakeholders begin, and how do we fix it?
How do we build counter-narratives to the fear of losing jobs in the communications and education fields?
How do we make professional discussions on AI use more diverse, where everyone’s voice is acknowledged?
What makes AI a medium, and how is AI different from other media?

References

Almuhanna, M. A. (2025). Teachers’ perspectives on integrating AI-powered technologies in K-12 education for creating customized learning materials and resources. Education and Information Technologies, 30(8), 10343–10371. https://doi.org/10.1007/s10639-024-13257-y

Buch, K., McCullough, H., & Tamberelli, L. (2017). Understanding and responding to the unique needs and challenges facing adjunct faculty: A longitudinal study. International Journal of Learning, Teaching and Educational Research, 16 (10), 27–40.

Cacioppo, J. T., & Patrick, W. (2009). Loneliness: Human Nature and the Need for Social Connection. W. W. Norton & Company.

Chekwa, C. (2018). Don’t be left out: Fostering networking opportunities to reduce workplace isolation among ethnic employees in remote settings. Journal of Competitiveness Studies, 26 (3–4), 217–235.

Flinders, D. (1988). Teacher isolation and the new reform. Journal of Curriculum and Supervision, 4(10), 17-29.

Gazzaniga, M. S. (2008). Human: The science behind what makes us unique (1st ed). Ecco.

Kanuka, H., Jugdev, K., Heller, R., & West, D. (2008). The rise of the teleworker: False promises and responsive solutions. Higher Education, 56(2), 149–165. https://doi.org/10.1007/s10734-007-9095-z

Okulicz-Kozaryn, W., Artyukhov, A., & Artyukhova, N. (2026). Will AI replace us? Changing the university teacher role. Societies, 16(1), 32. https://doi.org/10.3390/soc16010032

Walters, W., Robinson, D., Barber, W., & Spicer, C. (2025). The physical, social, and professional isolation of physical education teachers. Curriculum Studies in Health and Physical Education, 16(1), 1–21. https://doi.org/10.1080/25742981.2024.2311116

Zinkula, J., & Mok, A. (2024, March 6). ChatGPT may be coming for our jobs. Here are the 10 roles that AI is most likely to replace. Business Insider. https://www.businessinsider.com/chatgpt-jobs-at-risk-replacement-artificial-intelligence-ai-labor-trends-2023-02