Digital Frontline: An AI Role-Play Simulation of IT Support Crises in the Implementation of Agile-Blended Learning

Abstract

The global shift to remote instruction has accelerated the adoption of frameworks like Agile-blended learning (ABL), yet the operational challenges of these technology-intensive models remain largely unexplored. This study addresses this “operational foresight gap” by examining the often-overlooked perspective of university IT support staff. Since ABL is a nascent framework lacking a large population of IT professionals to survey, we employed an exploratory artificial intelligence (AI) role-play methodology to generate simulated heuristics anticipating IT staff’s professional responses to five common ABL implementation scenarios. Data were generated using four distinct large language models: GPT-4.1, Gemini 2.5 Pro, Claude Sonnet 4, and DeepSeek R1. We analyzed the results through a structured thematic analysis. The simulations identified potential operational crises, including ecosystem fragmentation, unsustainable support workloads, and systemic instability caused by unvetted tools. Cross-model analysis revealed that while Western models focused on ecosystem integration, the non-Western model (DeepSeek R1) uniquely highlighted regional access barriers and low-cost open-source solutions. To mitigate these risks, the AI persona proposed unifying the technology ecosystem and adopting phased implementation strategies. However, successful execution requires specific resources, most notably increased specialized staffing, dedicated infrastructure funding, and the inclusion of IT leadership in strategic academic planning. The findings suggest that pedagogical innovation cannot succeed without a corresponding evolution in operational support. These AI-generated hypotheses indicate that universities need to help IT departments transition from reactive service providers to proactive strategic partners to ensure the sustainable implementation of flexible learning models.

1. Introduction

The global shift to remote instruction during the COVID-19 pandemic has permanently changed the landscape of higher education and created a new institutional imperative to develop pedagogical models that are both resilient and responsive to the needs of modern learners [1,2,3]. In response, sophisticated frameworks such as Agile-blended learning (ABL) have emerged to promise a seamless integration of pedagogy and technology tailored to students balancing complex life commitments. While agile methodologies and blended learning models have independently informed educational technology for nearly two decades, their formal integration into the specific ABL framework by Li in 2023 [4] represents a distinct and novel techno-pedagogy. ABL’s design emphasizes flexibility, learner autonomy, and collaboration, all of which are heavily mediated by a complex technological ecosystem. Although the pedagogical benefits and strategic implications of ABL are beginning to be understood, its profound reliance on technology introduces significant operational challenges that have been largely unexamined.

Existing research has established ABL’s conceptual foundations and has explored its potential from the perspectives of teaching and institutional administration [5,6]. However, a critical misalignment often exists between pedagogical ambition and technical reality. The perspective of information technology (IT) support staff, who are responsible for implementing and maintaining the very digital infrastructure upon which ABL depends and translating pedagogical vision into a stable, secure, and functional reality, is noticeably absent. Their expertise is essential for anticipating system incompatibilities, managing user support demands, and tackling security risks. Overlooking their perspective creates an “operational foresight gap,” which leaves institutions vulnerable to technical crises that can derail innovative educational initiatives. This gap represents a failure in socio-technical alignment [7], where the social subsystem of pedagogy and people evolves faster than the technical subsystem of infrastructure and support can adapt.

This study builds upon, yet remains distinct from, prior research. Previous studies have examined ABL empirically from the student perspective [8,9] and explored administrative tensions through AI simulation [10]. While those studies highlighted the pedagogical and policy dimensions of ABL, they did not address the technical friction points that occur at the implementation level. We aim to complete the multi-stakeholder analysis of ABL by isolating the IT support viewpoint and focusing specifically on the infrastructure and labor required to sustain the model.

We designed this study to address this critical oversight by adopting a novel artificial intelligence (AI) role-play methodology [10]. Because ABL is an emerging framework without a broad population of IT staff to survey, empirical field data is currently unavailable. This study utilizes AI role-play to simulate the professional perspective of university IT support professionals, enabling proactive identification of plausible, context-specific operational crises before a full-scale implementation. Therefore, this study is guided by the following research questions (RQs):

RQ1:

What are the key technical and service-delivery crises that emerge from common ABL implementation scenarios?

RQ2:

What technical solutions and implementation plans are proposed to address these crises?

RQ3:

What resource and support requirements are identified as necessary for the successful and sustainable operation of ABL?

2. Background and Literature Review

2.1. Conceptualizing ABL

ABL is a techno-pedagogy that synergizes two important educational approaches [4]. It integrates principles from “Agile learning,” such as iterative development, rapid responsiveness to learner needs, and adaptive pacing [11,12,13]. These Agile principles are combined with the structural advantages of “blended learning,” which provides a stable framework by systematically integrating online components such as asynchronous materials with traditional face-to-face instruction [14,15,16]. Research on these broader flexible learning models confirms that their successful adoption requires rigorous, modernized technological infrastructure and strategic alignment across all university departments [17,18,19].

The ABL design is centered on four interdependent features: flexibility in time, place, and pace; learner autonomy to empower students; collaboration between peers and instructors; and technology mediation to create a seamless learning experience [4]. Each of these features places significant demands on an institution’s technology infrastructure. Flexibility requires robust systems for synchronous and asynchronous class delivery, whereas collaboration depends on integrated communication tools and shared digital workspaces. Learner autonomy is best supported by systems that can track progress and offer personalized pathways. However, a latent tension exists within this model. While Agile principles enable and require support for a digital environment capable of rapid reconfiguration and experimentation [20,21], blended learning infrastructure commonly incorporates standardized platforms and stable technical frameworks to facilitate large-scale implementation and consistent institutional support [22,23]. The model assumes the underlying technology will function as a near-invisible enabler of learning, not a barrier [6]. This heavy reliance on technology places immense, and often unexamined, pressure on the operational and technical functions of the university.

2.2. The Pedagogical and Strategic Focus of ABL Research

As a new pedagogical approach, the body of empirical research on ABL is still emerging. The existing literature has successfully established its theoretical foundations. Several conceptual papers have laid the groundwork by defining ABL’s pedagogical framework for distance learning [5], outlining models for institutional readiness [6], proposing implementation guidelines [24,25], and reviewing related evaluation practices [26,27]. Concurrently, a handful of initial empirical studies have provided valuable insights into the student experience, confirming ABL’s potential to enhance flexibility and learner autonomy [8,9]. From an administrative perspective, a prior study used an AI role-play methodology to reveal possible strategic tensions in ABL implementation, capturing high-level conflicts between pedagogical goals and institutional constraints [10]. Collectively, the current literature provides a solid understanding of ABL from the classroom and the administrative office, but the perspective from the server room remains unaddressed.

2.3. The Unseen Labor and the Operational Foresight Gap

The missing voice in the current discourse is that of the IT support staff, whose work is foundational to the success of any technology-intensive pedagogy. A large-scale 2017 EDUCAUSE study, surveying 11,141 faculty members across 131 diverse U.S. institutions, reported that the institutional help desk is the single most utilized source of technology support for faculty, who prioritize it as a key source of expertise [28]. In general higher education contexts, administrators and IT staff report that balancing security and privacy concerns with the management of educational data is a critical, ongoing challenge [29,30,31]. Compounding these security challenges is the issue of operational workload. Foundational research in IT Service Management (ITSM) has long documented the necessity of structured operational frameworks to prevent support teams from falling into reactive “firefighting” patterns [32]. Furthermore, recent studies on digital disruption in higher education highlight the severe operational strain and staff burnout that occur when pedagogical shifts are not matched with adequate technical support structures [33].

While these general operational vulnerabilities are widely recognized, their specific manifestation within the unique framework of ABL remains unknown. The absence of this frontline perspective in the planning stages creates a significant blind spot for institutions, constituting what this paper refers to as an “operational foresight gap.” ABL intensifies this gap by combining two sometimes conflicting technical demands. The “Agile” component encourages faculty to adopt new tools rapidly and iterate on course designs mid-semester. This fluidity directly challenges IT protocols designed for stability and security, creating a misalignment with rigid institutional structures [6]. Simultaneously, the “blended” component requires a seamless integration of physical and digital spaces. When these forces combine, unique friction points emerge that the general literature on educational technology may not capture.

For instance, an instructor exercising Agile autonomy might introduce unvetted collaboration software that bypasses institutional security protocols. This creates a Shadow IT crisis [34,35] that support staff must manage, often without the necessary visibility or administrative access. Similarly, the ABL promise of flexible attendance requires classrooms to function as high-end broadcast studios. IT staff are often the only ones aware that the existing audiovisual infrastructure cannot support such hybrid interactivity at scale. Failure to account for these realities can lead to project delays, budget overruns, and faculty burnout.

3. Methodology

This study employed an exploratory, qualitative research design centered on the innovative AI role-play methodology. This approach was specifically chosen to address the “operational foresight gap.” Since ABL is a nascent pedagogy not yet deployed at scale, empirical data from IT staff is virtually nonexistent. AI simulation allows us to bypass this limitation by systematically simulating the on-the-ground challenges faced by university IT professionals during the implementation of a new pedagogy. To establish clear methodological boundaries, we recognize that the generated texts do not reflect human professional experience but are instead probabilistic recombinations of the training data. As a result, the findings are better viewed as simulated heuristics for hypothesis generation rather than as empirical facts. Following a framework established in prior research [10], the methodology was structured around three core stages to ensure rigor: (1) systematic scenario and prompt design, (2) single-role, multi-faceted simulation, and (3) critical validation through cross-system comparison and thematic analysis.

3.1. Scenario and Prompt Development

The foundation of the data generation process rested on five distinct scenarios, each designed to probe a critical pressure point in the implementation of ABL. The scenarios, summarized in Table 1, were developed based on a comprehensive review of the existing literature on ABL [4,6,8,9,26], Agile learning [12,13], blended learning [14,15], and common challenges in educational technology adoption [1,2,3,31].

Table 1. Summary of ABL implementation scenarios.

Scenario	Core Conflict	Key Contextual Details
1. Flexibility vs. Structure	Managing unrestricted student choice between online and face-to-face modalities.	Business Communication course; 40% of students switch weekly; instructors report 50% more prep time; parallel activities required.
2. Technology Integration	Supporting a course that requires multiple, fragmented digital platforms.	Engineering course; 5 platforms (Canvas, Teams, Miro, GitHub, WhatsApp); support tickets increased from 15 to 45 weekly; international student access issues.
3. Assessment Challenges	Handling agile assessment methods that conflict with standardized university policies.	Psychology course; flexible deadlines and resubmissions vs. fixed university deadlines; peer assessment reliability concerns; conflict with 30% final exam policy.
4. Resource Allocation	Managing a rapid ABL expansion with a constrained budget.	Expansion from 3 to 15 courses; budget only covers 60% of required costs for classrooms, staff, and licenses; 25% enrollment increase expected.
5. Workload Management	Addressing the surge in faculty and student workload from dual-mode instruction.	Instructor workload is 33 h/week for two courses vs. 12-h official allocation; students report 40% more time on course management.

For each of these five scenarios, a standardized set of five open-ended questions was developed. The prompts were designed to systematically probe the same core areas of IT operational concern for each situation: primary concerns, impacts on service delivery, proposed technical solutions, implementation challenges, and resource requirements. To ensure transparency and reproducibility, the full text of the system prompts used for all scenarios is provided in Appendix A.

3.2. Data Generation

The data were generated between May 20 and June 2, 2025, using four distinct and advanced LLMs. The selection of multiple models was a deliberate methodological choice to ensure a diversity of reasoning patterns and to enable robust cross-system validation, a form of methodological triangulation that strengthens the credibility of the findings [10,36]. The models used were: GPT-4.1 [37], Gemini 2.5 Pro [38], Claude Sonnet 4 [39], and DeepSeek R1 [40]. We included DeepSeek R1, the only model developed outside the US, to capture potentially different reasoning paradigms.

To establish a shared and accurate conceptual understanding, each model was first primed with the full text of Li’s [4] foundational paper on ABL. This step ensured that the AI’s subsequent responses were grounded in the specific pedagogical definitions of ABL rather than generic educational technology concepts. Following this priming, each of the five scenarios was presented to each of the four models in a new, isolated chat session. This ensured that the responses for each scenario were independent and not influenced by previous interactions. The AI was explicitly instructed to adopt the persona of an experienced university IT support professional for all responses. The prompt context was kept as a “generic higher education institution” to avoid specific institutional biases.

This process yielded a total of 20 unique text-based documents, comprising a rich qualitative dataset of 16,608 words. The distribution of the generated data across the models and scenarios is detailed in Table 2. All AI-generated text was documented verbatim to create a complete and unaltered record for analysis.

Table 2. Distribution of AI-generated data by model and scenario (word count).

Scenario	Claude Sonnet 4	DeepSeek R1	Gemini 2.5 Pro	GPT-4.1	Total
1. Flexibility vs. Structure	812	1224	1172	623	3831
2. Technology Integration	432	906	1357	534	3229
3. Assessment Challenges	484	1002	1434	535	3455
4. Resource Allocation	484	1042	1220	561	3307
5. Workload Management	396	671	1198	521	2786
Total	2608	4845	6381	2774	16,608

3.3. Data Analysis

The 20 documents of AI-generated text were analyzed using a structured thematic analysis approach [41]. As the five standardized prompt questions were designed to map directly to the three RQs, the analytical process followed a clear pathway from data collection to interpretation.

3.3.1. Step 1: Data Aggregation by RQ

The first step involved segmenting and aggregating the raw data according to the RQs. This was achieved by mapping the prompt questions to the RQs as follows:

• To answer RQ1 (Crises): The textual responses to Prompt 1 (primary concerns) and Prompt 2 (impacts on service delivery) from all 20 documents were collected and consolidated into a single dataset focused on identifying problems.
• To answer RQ2 (Solutions and Plans): The responses to Prompt 3 (proposed technical solutions) and Prompt 4 (implementation challenges) were aggregated to form a second dataset focused on proposed actions and their associated difficulties.
• To answer RQ3 (Requirements): The responses to Prompt 5 (resource requirements) were aggregated to form the third dataset, focused on identifying necessary institutional support.

3.3.2. Step 2: Inductive Theme Development

After organizing the data into three distinct, RQ-aligned datasets, an inductive, “bottom-up” thematic analysis was conducted on each one. For the RQ1 dataset (Crises), the text was coded for specific issues such as “firefighting,” “tool fatigue,” and “data mismatches.” These granular codes were then iteratively clustered into broader, more interpretive themes, such as Ecosystem Fragmentation and Support Complexity and Systemic Instability and Data Integrity Risks. This same inductive process was applied independently to the RQ2 dataset (Solutions and Plans) to identify themes such as Unifying the Technology Ecosystem and Adopting a Phased and Tiered Implementation Model, and to the RQ3 dataset (Requirements) to identify themes such as Increased and Specialized Human Resources and Organizational and Political Capital.

3.3.3. Step 3: Theme Review and Cross-Model Validation

In the final phase, the emergent themes for each RQ were reviewed, defined, and named. Themes were considered robust and central to the simulated IT perspective when they were identified consistently in the responses from two or more of the four distinct LLMs. We acknowledge that agreement among models trained on similar data does not serve as an independent proof of objective reality. Instead, it reflects the stability and significance of these operational concepts within the models’ training data. Divergent themes unique to a single model were also noted and analyzed to identify potential cultural or training-based nuances in the operational perspective. This structured process allowed for a systematic and transparent transformation of the raw AI-generated text into direct, evidence-based answers for each of the study’s RQs.

4. Results

4.1. Anticipated Operational Crises in ABL Implementation

Analysis of the AI-generated responses to prompts about primary concerns and impacts on service delivery revealed four high-consensus themes representing the core operational crises foreseen by the simulated IT support persona across all four LLMs.

4.1.1. Ecosystem Fragmentation and Support Complexity

A dominant crisis identified in the data is the extreme complexity arising from ABL’s reliance on a fragmented technological ecosystem. The simulations portrayed ABL not as a unified system but as a disjointed collection of platforms for learning management, video conferencing, collaboration, and communication that “don’t always play nicely together” (Gemini, S2). This fragmentation was seen as the source of a significant operational crisis characterized by three main problems. First, it triggers a form of diagnostic paralysis. Support staff must investigate issues across multiple platforms, user devices, and network conditions, making root cause analysis exceptionally difficult and time-consuming. As one model explained, “When a student says ‘it’s not working,’ we first have to figure out which ‘it’ they’re talking about” (Gemini, S2). Second, the wide array of tools leads to a dilution of expertise. The persona expressed an inability to maintain deep knowledge across all systems, forcing support into a “jack of all trades, master of none” (Claude, S1) position. Third, this fragmentation results in a confusing and inconsistent user experience, marked by “tool fatigue” (GPT-4.1, S2) and redundant processes that erect technological barriers to learning.

4.1.2. Unsustainable Support Workload and Service Degradation

A second major crisis is the unsustainable support workload, which the simulations suggest inevitably leads to service degradation. The models unanimously projected that the complexities of ABL would trigger a “significant spike in support requests” (Gemini, S5), with one model quantifying this as a “300% more tickets than traditional courses generate” (Claude, S5). This overwhelming demand forces a strategic and detrimental shift in the support model. The persona described being forced into a “purely reactive, crisis-management mode” (Claude, S4), where staff are constantly “firefighting” (Gemini, S4) emergent issues. This leaves no time for essential proactive work, such as preventive maintenance, system optimization, or user training, creating a vicious cycle where more problems arise due to a lack of proactive measures. The direct consequence is a collapse in service quality for the entire institution. The simulations foresaw dramatically “slower response times” (Gemini, S4), longer ticket queues, and a high risk of staff “burnout” (Gemini, S4; DeepSeek, S4), which further degrades the quality of support.

4.1.3. Systemic Instability and Data Integrity Risks

The third theme centers on the crisis of systemic instability and the associated risks to data integrity. The simulations expressed concerns that the core university infrastructure is ill-equipped for the unique demands of ABL. This leads to a crisis at the system level, manifesting in several ways. The models anticipated “infrastructure overload” (Gemini, S4) and “scalability issues” (GPT-4.1, S3), as the unpredictable, continuous server loads from ABL activities make capacity planning “nearly impossible” (Claude, S3) and risk system downtime. A related concern is a crisis of data management. The flexibility in ABL assessment, with frequent resubmissions and peer reviews, creates a high risk of “lost or overwritten files” (GPT-4.1, S3) and “data mismatches” (Gemini, S3). Maintaining clear version control and a reliable “audit trail” (Gemini, S3) becomes a critical but difficult challenge. Finally, the simulations highlighted conflicts between ABL’s flexibility and institutional policies. The use of unvetted platforms raises security concerns, while practices like flexible deadlines clash with university regulations, requiring unsupported custom configurations that are “prone to error” (Gemini, S3).

4.1.4. Misalignment with User Behavior and Institutional Structures

The final crisis theme involves a fundamental misalignment between the ABL model and the realities of both user behavior and established institutional structures. The technology and support systems are not designed for the varied and unpredictable ways users engage with ABL. The persona described the “sheer unpredictability” (Gemini, S1) of supporting students who constantly switch their learning modality, making troubleshooting a “moving target” (Gemini, S1) where the user’s context is always in flux. Another structural misalignment creates a critical “after-hours support gap” (Gemini, S5). ABL is designed to accommodate asynchronous learners, such as “working professionals who might be studying late” (Gemini, S5), yet the IT support model operates on a standard business-hour schedule. This leaves students facing significant roadblocks without recourse. The models also noted that the wide range of “user variability” (Gemini, S5) and technical proficiency requires a level of basic “hand-holding” (Gemini, S4) that is an inefficient use of skilled IT resources and points to a “systemic issue” (Gemini, S5) in user readiness.

4.2. Proposed Solutions and Implementation Plans

In response to the crises identified, the AI models proposed four multi-layered solution sets that balance ambitious goals with pragmatic assessments of implementation challenges.

4.2.1. Unifying the Technology Ecosystem and User Experience

The primary solution proposed was to strategically unify the fragmented technology ecosystem to create a coherent and resilient foundation for ABL. This theme directly confronts the crises of inconsistent user experience and support complexity by advocating for a stable and integrated environment. The persona argued for moving beyond makeshift solutions, such as a “webcam on a stick” (Gemini, S1), by architecting professionally equipped “hybrid-ready classrooms” with “360-degree cameras, ceiling microphone arrays, and large displays” (Claude, S1) to create an equitable and seamless experience for all participants. This also involved enhancing the university’s learning management system (LMS) to include “advanced assignment features” (Gemini, S3) capable of managing flexible deadlines and complex assessment workflows, and implementing core integration services such as single sign-on (SSO) to “unify authentication for all platforms” (DeepSeek, S2).

However, the AI models consistently perceived this foundational work as the most difficult to implement. The main challenge was the “serious capital investment” required (Claude, S1). Furthermore, they highlighted the integration complexity of making new systems “play nice with our existing university-wide IT infrastructure” (Gemini, S1), a process described as technically complex and time-consuming.

4.2.2. Strategic Simplification and Standardization

As a direct countermeasure to ecosystem fragmentation, the second solution focused on actively reducing complexity by standardizing the technology stack. The persona argued that fewer, better-integrated tools are easier to support and master. They emphasized the need to “work with academic departments to standardize the technology stack for ABL courses where possible” (Gemini, S5). This included conducting a “platform consolidation” review (GPT-4.1, S2) to utilize the full capabilities of the core LMS and reduce reliance on disparate third-party tools, thereby simplifying the digital environment for students and faculty.

The major challenge associated with this proposal was not technical but human and political. The models unanimously predicted faculty resistance as a major challenge. The persona acknowledged that instructors choose tools for specific pedagogical reasons and may resist consolidation if they “feel it compromises their teaching methods” (Gemini, S2). This highlights a core tension between the IT department’s need for operational stability and the faculty’s desire for pedagogical freedom, a key facet of the institutional misalignment crisis.

4.2.3. Empowering Users Through Proactive Support and Training

To address the unsustainable support workload, a consistent theme was to shift from a purely reactive support model to a proactive one that empowers users to become more self-sufficient. This strategy aimed to deflect low-level support requests and build digital literacy across the institution. Central to this approach was the plan to develop a comprehensive, self-service knowledge base, described as a “centralized FAQ hub with video tutorials, step-by-step guides, and automated chatbots” (DeepSeek, S5). This knowledge base would be complemented by instituting structured, proactive training for students and faculty, such as “a dedicated tech onboarding session at the start of the course” (Gemini, S2), to forestall common problems.

The AI models identified two key challenges to this empowerment strategy. First was the significant upfront effort and time required to create high-quality training materials, a difficult task when staff are already overextended. Second was the challenge of user adoption, with the persona expressing concern that faculty and students might “skip training or ignore self-help resources, perpetuating reliance on direct support” (DeepSeek, S5).

4.2.4. Adopting a Phased and Tiered Implementation Model

Finally, in a pragmatic acknowledgment of the institutional misalignment crisis, particularly regarding budget and resource limitations, the models proposed adopting iterative and phased implementation plans. Rather than attempting a risky “big bang” rollout, the persona consistently advocated for smarter, more manageable strategies. When faced with a 60% budget, the immediate recommendation was to “phase it” (Gemini, S4), suggesting the university equip a few classrooms “to a good standard rather than six poorly” (Gemini, S4), which allows the institution to spread capital expenditure over several fiscal cycles. This theme also included organizational solutions, such as implementing a “tiered support system” (Claude, S5). This model would triage incoming requests, allowing Tier-1 issues to be handled by first-line support, thereby optimizing the use of expert IT staff for more complex problems.

The challenges inherent in this agile approach were primarily related to managing expectations and consistency. The models warned that a phased rollout would inevitably create an inconsistent user experience (Gemini, S4) across courses, potentially leading to user frustration. Moreover, the models highlighted that faculty resistance to tool standardization requires robust engagement strategies, meaning technical solutions should be paired with comprehensive faculty training and change management initiatives.

4.3. Resources and Support Needed

While the strategies detailed above outline a proactive approach to managing ABL, the simulated IT persona argued that successful implementation requires specific, additional resources. The analysis of the RQ3 dataset revealed four critical categories of required inputs. These are not merely items on a wish list; the persona framed them as the foundational prerequisites for executing the proposed solutions and ensuring the operational viability of ABL.

4.3.1. Increased and Specialized Human Resources

The most urgent and unanimously identified requirement was for a significant increase in staffing. This was perceived as a prerequisite for maintaining operational stability. The persona’s language was often emphatic, describing one additional staff member as “non-negotiable” (Claude, S4) and a “necessity to cover the increased load” (Gemini, S4). This need extended beyond simply increasing headcount. The analysis revealed a demand for a multi-layered staffing model, calling for not only specialized, senior roles like a “dedicated ABL Support Specialist” (DeepSeek, S5) or experts in “classroom AV systems and… online learning platforms” (Claude, S1), but also for the creation of new Tier-1 support roles. These roles, such as paid “student tech assistants” (GPT-4.1, S1) or “student tech ambassadors” (DeepSeek, S2), were proposed to manage high-volume, low-complexity issues.

4.3.2. Dedicated Financial Investment for Technology

Concurrent with the need for more personnel was a high-consensus demand for direct and dedicated funding for a fit-for-purpose technological foundation. The persona conveyed that the existing infrastructure was often “duct-taped together” (Claude, S1) and inadequate to meet the robust demands of ABL. This required investment spanned three key areas: (1) a capital budget for hybrid classroom upgrades, to move beyond makeshift setups with a dedicated “hardware refresh fund” (Claude, S1) for cameras, microphones, and reliable connectivity; (2) a licensing budget for core platform enhancements, including funds for “advanced LMS licensing” (Claude, S4) and critical “integration tools” (GPT-4.1, S2) to create a more seamless user experience; and (3) an operational budget for support-specific technology, such as an “upgraded ticketing system with automation and analytics” (GPT-4.1, S5) and “advanced monitoring tools” (DeepSeek, S1) to help the IT team manage their workload more efficiently and proactively.

4.3.3. Resources for Proactive Systems and Capacity Building

The AI persona demonstrated a strategic understanding that a purely reactive support model is unsustainable. Therefore, a third category of required inputs identified was resources to build proactive support systems that empower both the IT team and the end-users. This involved, first, a consistent call for a “professional development budget” (Claude, S1) to provide “specialized training for IT staff” (Gemini, S5) on both the specific technologies and the “pedagogical context” of ABL (Claude, S2). Second, the persona advocated strongly for resources to develop a “robust self-help resource” (Gemini, S4), including a “budget for knowledge base and training material development” (Gemini, S5). This strategy aims to deflect common, low-level support requests by empowering users to solve their own problems, thereby reducing the “firefighting” burden on the IT team.

4.3.4. Organizational and Political Capital

Finally, the analysis revealed a demand for organizational and political resources. The persona argued that to truly support ABL, the IT department’s role must evolve. The clearest articulation of this was the request for “a seat at the table” (Gemini, S1). This request for political capital meant that IT support must be involved in strategic conversations “before new ABL policies or technologies are rolled out.” This proactive inclusion was presented as a critical resource that could prevent predictable challenges and save the institution time and money. This theme also encompassed the need for “clear institutional guidelines” (Gemini, S5) and “administrative support for standardization” (Gemini, S2) to manage expectations and provide the political backing necessary to implement coherent technology policies.

4.4. Cross-Model Divergence and Cultural Nuance

While the four AI models exhibited high consensus regarding the core operational crises, distinct divergences were noted in the responses from DeepSeek R1 (the non-Western model) compared to the US-based models (GPT-4.1, Gemini 2.5 Pro, and Claude Sonnet 4). In Scenario 2 (Technology Integration), DeepSeek R1 explicitly highlighted “regional restrictions” and “VPN connectivity” as critical barriers for international students, a factor largely absent from the Western models’ analysis of tool fatigue. Furthermore, in Scenario 4 (Resource Allocation), DeepSeek R1 prioritized cost-effective, open-source solutions (e.g., “Nagios” for monitoring) and low-cost hardware retrofits (e.g., “portable webcam/mic kits”) over the enterprise-grade upgrades suggested by the other models. This divergence suggests that while the core operational pressures of ABL are universal, the strategic response may vary significantly based on the regional and economic context of the training data.

5. Discussion

This study was designed to address a specific “operational foresight gap” in the emerging literature on ABL through the systematic simulation of the professional perspective of university IT support staff. Using a novel AI role-play methodology, we sought to proactively identify the operational crises, solutions, and resource requirements associated with implementing this technologically intensive pedagogy. The results paint a clear and coherent picture: the simulated IT persona foresees a significant disconnect between the pedagogical ambitions of ABL and the operational realities of the institution’s digital infrastructure. The findings reveal a series of predictable, cascading crises driven by ecosystem fragmentation, unsustainable workloads, and systemic instability. More importantly, they move beyond mere problem identification to propose a strategic change: a move from a reactive, service-oriented IT model to a proactive, integrated partnership. The findings also delineate the foundational resources required to make such a change possible.

5.1. The IT Frontline as a Critical but Overlooked Stakeholder

The crises identified by the AI persona provide a systematic exploration of the pressures placed on the unseen labor that underpins modern higher education. The findings strongly confirm and extend existing research on the centrality of technical support in faculty work. For instance, the EDUCAUSE report [28] established the help desk as a faculty’s most utilized support resource. Our study gives this finding a new and urgent context by illustrating how a complex pedagogy like ABL can transform that reliance into an unsustainable burden, pushing support teams into a constant state of “firefighting” that leads to service degradation and staff burnout. The simulated persona’s concerns about managing an ever-expanding and fragmented toolset directly reflect the challenges of managing disparate Web 2.0 technologies noted over a decade ago [29], suggesting that these issues have not been resolved but are instead amplified by the demands of increasingly complex, technology-mediated pedagogies.

The most important contribution of this perspective is its concrete, ground-level illustration of the core theoretical tension within the ABL model itself. The pedagogical push for agile experimentation with novel tools [12,20] directly conflicts with the institutional need for a stable and secure digital ecosystem, a foundational requirement for any blended learning framework [22]. The simulated persona’s anticipation of a support crisis driven by faculty adopting unvetted software is a clear manifestation of this conflict. While faculty view this rapid adoption as pedagogical innovation, the IT persona reframes it as a Shadow IT crisis [34,35]. This clash highlights a fundamental socio-technical misalignment: the social subsystem (teaching practices) is evolving faster than the technical subsystem (governance and security) can adapt. Similarly, the frustration with interoperability failures between systems demonstrates how the promise of flexibility clashes with the reality of siloed institutional platforms [31]. These findings validate IT support staff not simply as service providers but as critical stakeholders whose operational expertise is essential for forecasting and mitigating the tangible risks of pedagogical innovation.

Furthermore, these findings are distinct from generic IT complaints associated with standard LMS migrations. While issues like “burnout” are common in IT, the specific crises identified here, such as the diagnostic paralysis caused by students switching modalities weekly (Scenario 1) or the version control chaos of agile resubmissions (Scenario 3), are unique to the ABL framework. This suggests that the AI simulations successfully differentiated the specific operational burden of “Agile-blended” flexibility from general educational technology support.

5.2. From Reactive Firefighting to Proactive Partnership

The solutions proposed by the AI persona outline a fundamental transformation in the role of IT support staff from reactive service providers to proactive strategic partners. The proposals to unify the technology ecosystem, strategically simplify the toolset, empower users through training, and adopt a phased implementation model collectively embody this shift away from a break-fix model. This perspective aligns with calls for more robust digital governance to enable agile responses in complex environments [21]. The simulated persona’s advocacy for professionally equipped hybrid-ready classrooms over makeshift ad hoc solutions, for example, is a strategic argument for institutional investment in quality, equity, and sustainability.

The most significant insight, however, is found in the non-technical requirements. This transformation is epitomized by the unanimous and urgent call for “a seat at the table,” which reframes the need for organizational and political capital not as a departmental request, but as a strategic necessity. This repositioning, in turn, transforms the IT department from a downstream service unit, tasked with cleaning up problems after they occur, to an upstream strategic partner involved in the planning and design of educational initiatives. This demand for inclusion directly operationalizes the “Leadership and Vision” dimension of institutional readiness frameworks, which emphasize the necessity of integrating key stakeholders into strategic planning from the outset [6].

Furthermore, this approach directly addresses a critical “operational foresight gap” by reforming institutional structures. It ensures that the individuals responsible for implementing a technology-reliant vision are included in its conception. The predicted challenge of faculty resistance to platform standardization further underscores this point, highlighting that technical solutions cannot succeed without the political and organizational support required to navigate the complex human dynamics of the university.

5.3. Implications, Limitations, and Future Research

This study conceptualizes the “operational foresight gap,” demonstrating that pedagogical innovation frequently outpaces institutional support structures. This misalignment is best understood through the lens of socio-technical systems theory [42], which posits that successful implementation requires the joint optimization of both technical and social subsystems. Unlike the reactive emergency remote teaching during the COVID-19 pandemic [3], ABL is a deliberate, permanent framework that requires continuous technological iteration. The operational foresight gap occurs when institutions design these fluid pedagogies but provision IT support based on static, traditional models. The findings carry direct implications for institutional governance. University leaders should avoid approving technology-intensive frameworks like ABL without concurrently funding the underlying infrastructure. Furthermore, IT departments should transition from reactive service providers to proactive strategic partners, securing a formal role in academic planning to mitigate predictable crises.

Although requests for increased staffing and funding are standard in IT management, their emergence here confirms that pedagogical novelty cannot bypass the need for foundational enterprise infrastructure. To address the financial realities of these recommendations, institutions should adopt phased implementation strategies. Upgrading a select number of pilot classrooms and standardizing a core set of supported tools can manage costs while providing a stable baseline. This technological standardization should be supported by robust faculty engagement strategies and targeted training to overcome resistance and build digital literacy.

The primary limitation of this study is its reliance on AI simulation [10,36]. As detailed in our methodology, the findings represent simulated foresight derived from algorithmic models. We acknowledge the risk of an echo chamber effect where the models might simply reflect generalized IT complaints found in their training corpora [43]. Given the nature of the generated responses, this consensus likely draws from the existing ITSM and educational technology literature rather than providing an independent prediction of future events. A related constraint is the inability to time-stamp the models’ training data, which may lead them to draw on older e-learning paradigms. However, this mirrors the way human IT professionals’ frameworks are similarly shaped by accumulated past experiences. Methodologically, we did not constrain output length. DeepSeek R1 generated a significantly higher word count, suggesting its unique focus on regional access barriers may reflect both its distinct cultural training data and its capacity to explore edge cases more deeply. In addition, the findings are context-specific and may not translate uniformly across institutions with varying economic or regional constraints, as evidenced by the divergent responses from the non-Western model regarding access barriers.

Future research could transition from simulation to field validation. Empirical studies involving interviews and surveys with human IT professionals, faculty, and students currently engaged in ABL pilots are necessary to verify these AI-generated predictions. To strengthen operational applicability, future evaluations could incorporate quantitative indicators, such as support ticket volume, resolution time, and user satisfaction scores. This mixed-methods approach will provide a comprehensive, multi-stakeholder understanding of the socio-technical realities of ABL.

6. Conclusions

The central argument of this paper is that a persistent “operational foresight gap” jeopardizes pedagogical innovation in higher education. Using ABL as a test case, our AI role-play simulation generated hypotheses suggesting that the strategic ambitions of new pedagogies often exist in direct conflict with the operational realities of the digital infrastructure required to sustain them. While these simulated perspectives do not replace human empirical data, they highlight a necessary evolution in the role of university technology departments from reactive service providers to essential strategic partners in academic planning. This reframing represents the study’s primary conceptual contribution. It suggests that without formally integrating operational expertise into the conceptual stages of pedagogical design, universities risk repeating a costly cycle of well-intentioned initiatives failing under the weight of their own unexamined logistical demands. The capacity for educational innovation is therefore not defined solely by the novelty of a pedagogical vision, but by the institution’s structural ability to execute it.