The review followed the PRISMA-ScR reporting guideline and searched Scopus, Web of Science, ERIC, the British Education Index and Education Source for English- and Spanish-language higher-education studies published from 1 January 2022 to 6 June 2026. Citation searching was added, and studies qualified if they supplied original empirical data or systematically documented, first-hand evidence of a practice already enacted with students. Twenty-eight studies met the empirical-data criterion and three entered through the practice-based route. Searches produced 21,720 records; after duplicate and retraction checks, 15,179 records were screened at title-and-abstract level and 3,003 at full-text level. The final corpus contained 31 studies, all published between 2023 and 2026, with 77.4% appearing in 2025 or 2026. Most concerned mixed-disciplinary settings, undergraduate or postgraduate coursework, or other academic-writing contexts.

From 29 contributing studies, the researchers charted 192 practice-level records describing particular uses of generative AI. Of these, 107, or 55.7%, were classified as responsible, 42 as contested and 43 as problematic. These proportions describe the extracted material and should not be interpreted as population prevalence, because studies that documented more practices contributed more records. The classification depended strongly on where AI entered the writing process. Language polishing and editing were generally treated as responsible when students retained control and evaluated suggestions. Brainstorming could also be acceptable when its output served as a prompt for further thought, but idea generation became problematic when formulating the research question was itself the assessed skill. AI-generated references and summaries were considered problematic if students treated them as authoritative evidence without checking the original publications, while using generated material as a draft or model was often contested because acceptability depended on how much of the student’s contribution remained identifiable.

Three of the six principles concerned decisions made directly during writing. The first, targeting learning objectives, asks whether assistance supports or displaces the capability an assignment is intended to develop. The same tool could therefore be acceptable in an advanced course but not in a beginners’ writing class. The second principle, preserving human agency, distinguishes a scaffold from a replacement. Responsible uses exposed students to possible ideas, revisions or feedback, after which they selected, rejected, adapted and explained the material. Contested cases involved uncertain participation, while problematic cases delegated substantive reasoning or final judgement to the system. The third principle requires verification. Definitions, literature suggestions, quotations, citations and factual claims generated by AI were treated as starting points only when students located and checked the originals. Because hallucinations, bias and tool limitations remain possible, students remain responsible for the accuracy and integrity of the final text.

The remaining principles concern accountability and protection. Transparency and accountability meant acknowledging assistance and following relevant course, institutional or journal requirements, although disagreement remained over whether minor uses such as grammar correction or limited brainstorming required disclosure. Protecting authorship and human voice focuses on ownership, originality and personal expression, especially in multilingual writing: AI-supported translation or polishing may reduce linguistic barriers, yet it can also displace culturally significant features or standardise a student’s voice. Privacy, consent and fairness address the conditions of use rather than the text alone, including confidential manuscripts, sensitive information, undisclosed AI evaluation of student work, unequal access to paid platforms and biased outputs. Comparisons involving 16 jointly reported student and staff records found agreement that grammar correction and paraphrasing one’s own text could be responsible, while full-text generation and replacement of human judgement were problematic. Divergence clustered around the uncertain boundary between support and over-reliance.

The review also mapped 122 educational approach records from 29 studies. Pedagogical approaches were the largest group, with 69 records, followed by guidance with 19, assessment approaches with 12 and policy-linked guidance with eight; 14 concerned supporting measures such as infrastructure, monitoring, accessible resources and modelling. Common teaching methods included workshops, induction sessions, course-embedded AI literacy, guided practice and reflection. To operationalise verification, educators asked students to triangulate AI-suggested references through conventional searching, check facts and sources, and compare their summaries with AI-generated alternatives. To strengthen human agency, some instructors asked students to request lists of suggested corrections and apply the changes themselves, while others used generated outlines, feedback and images as objects for evaluation rather than finished prose. Approaches focused on learning objectives by defining permitted and prohibited uses separately for each writing task.

Implementation was mixed. Of the 122 approach records, 52 were proposed without reported implementation, 19 had been implemented without evaluation and 51 had been implemented with some form of evaluation. The evaluated group included 36 pedagogical approaches, nine guidance or policy-linked approaches, three assessment approaches and three supporting measures. Assessment redesign proposed making the writing process visible through planning notes, version histories, prompt records or reflective accounts, and pairing written work with oral examination. Module-level guidance was often tied to learning outcomes and student developmental level, while institutional proposals included collaborative policy development, staff training, regular review and provision of a common platform intended to support equal access. The review notes, however, that an evaluation of an overall programme does not establish that every component or every principle within it was tested.

Evaluations were strongest for proximal outcomes such as students’ confidence, understanding of institutional expectations, awareness of AI limitations, prompting and reported verification behaviour. One study reported a small but statistically significant improvement in self-reported verification relative to a control group, with a partial eta squared of 0.01 and p below 0.05, but the associated reduction in trust in ChatGPT was not significant. Other studies used pre-post self-reports, post-class reflections, classroom artefacts and instructor observations. Direct evidence on writing performance was less common, appearing in rubric-based pre-post writing assessments and comparisons of legal-redrafting across classroom conditions. Qualitative findings indicated that students could identify weaknesses in AI prose and articulate context-specific boundaries. These results show that instruction can accompany changes in reported understanding and practice, but they do not establish a general effect on writing quality, academic integrity or long-term responsible use.

The clearest research gap is the missing chain between a stated boundary, an intervention designed around it and an evaluation of whether students follow it. Although 27 of the 31 studies contributed to both the practice and approach datasets, contributing records did not mean those elements had been linked in a tested programme. The synthesis therefore supports a division of responsibility: institutions could establish baseline expectations for verification, disclosure, accountability and data protection, while individual courses specify how AI assistance relates to assessed learning. Teaching designs could make prompts, AI outputs, successive drafts and revision decisions visible, allowing instructors to discuss whether assistance supports or replaces the intended skill. The authors also call for discipline-specific work, because evidence from mixed settings cannot show how a boundary changes between subjects, academic levels or assessments. Future studies should test decision-support tools derived from the six principles and measure disclosure accuracy, source checking, retained authorship and transfer across new tasks.

The review has important limitations. It excluded grey literature, preprints and conceptual or policy-only work, and Spanish-language retrieval depended on the English search terminology. Full-text screening and initial coding were primarily conducted by one reviewer, although co-authors later audited 119 original practice records and 79 approach records from 17 studies. No formal risk-of-bias appraisal was conducted, and the six principles emerged through interpretive synthesis and author discussion rather than a validated classification system. Record-level counting also gave greater numerical weight to studies with more detailed reporting. With those constraints in mind, the safest conclusion is not that generative AI has a proven place in academic writing, but that responsible use cannot be specified as a universal list of tools or prohibited functions. Its defensibility depends on the assignment’s purpose, the student’s intellectual contribution, the safeguards attached to use and whether responsibility remains with a person who can verify, explain and defend the final work.

Subject of Research: Responsible generative AI use in higher-education academic writing

Article Title: Responsible student use of generative AI in academic writing: a scoping review of boundaries, educational approaches, and evidence gaps

Article References: Tie, J., Gil Ortega, M., & Zhao, D. (2026). Responsible student use of generative AI in academic writing: a scoping review of boundaries, educational approaches, and evidence gaps. Frontiers in Education, 11, Article 1998449. https://doi.org/10.3389/feduc.2026.1998449

Image Credits: AI Generated

DOI: 10.3389/feduc.2026.1998449

Keywords: generative AI, academic writing, higher education, student learning, academic integrity, AI literacy, human agency, output verification, scoping review, assessment design, Responsible, student

The framework’s central insight rests on a distinction that cognitive load theory has refined over decades: not all mental effort is equal, and not all of it deserves to be automated. Intrinsic load refers to the inherent complexity of a task — the number of elements a writer must hold in mind simultaneously to make sense of an argument. Extraneous load, by contrast, is the friction that contributes nothing to understanding: hunting for a citation format, wrestling with a template, untangling a garbled bibliography. Germane processing is the deliberate allocation of working memory resources toward building schemas, the mental structures that make expertise possible. The paper’s argument is that AI tools are remarkably well suited to absorbing extraneous load and strikingly ill suited to absorbing germane processing, because germane processing is not a task that can be handed off at all. When a tool performs the work, the processing does not transfer to the writer; it simply does not occur.

This claim gains force from recent empirical findings on what happens when writers lean heavily on generative assistance. Studies documenting reduced neural engagement, weaker recall of one’s own writing, and a diminished sense of authorship among heavy AI users suggest that the costs of offloading are not merely pedagogical abstractions. They are measurable changes in how deeply a writer encodes their own ideas. Writing, as Janet Emig argued in 1977, is a distinct mode of learning rather than a record of thinking already completed. The act of struggling to render a messy idea in precise language is itself the mechanism by which the idea becomes clear to the writer. Remove the struggle, and the clarity never arrives.

The evidence on writing-to-learn adds an important nuance. A review of 48 school-based writing-to-learn programs by Bangert-Drowns and colleagues found a consistent positive effect on achievement, and a subsequent synthesis by Graham and colleagues reported comparable effects across 56 experiments in science, social studies, and mathematics. But what predicted a stronger effect is instructive for AI design: writing that included metacognitive prompts helped more, while assigning longer pieces helped less. The benefit of writing comes from the requirement that writers think about what they are doing, not from sheer output volume. This is precisely the condition that AI can silently remove. A student who prompts a chatbot to expand a paragraph has produced more words while doing less thinking — the exact inversion of what the writing-to-learn literature identifies as beneficial.

The nine-phase framework responds to this problem by refusing to treat research writing as a single undifferentiated activity. Instead, it maps the process from initial curiosity and inquiry through references and submission, and for each phase specifies three things: the dominant cognitive demand, the appropriate AI role ranging from minimal involvement to full collaboration, and design guidelines for how tools and interactions should behave. The logic is that the right AI role depends on the cognitive demand of the phase, not on the tool’s capabilities. In early phases, where the writer must generate questions and confront ambiguity, AI involvement should be minimal, because the struggle to formulate a researchable question is germane processing in its purest form. In later phases, where formatting conventions and citation management dominate, AI can operate at full collaboration, because those demands are largely extraneous and offloading them frees working memory for the thinking that matters.

This phase-aware approach is an application of responsive pedagogy to tool design. Responsive pedagogy holds that instruction should respond to the learner’s current understanding through recursive feedback dialog rather than proceeding on a fixed script. Transposed to AI-assisted writing, the principle becomes: the tool should respond to the cognitive demands of the task at hand, adapting its level of involvement as the writer moves through phases. A system that offers the same degree of assistance at every stage — drafting, organizing, synthesizing, polishing — is pedagogically blind, because it cannot distinguish between the friction that should be removed and the friction that is doing the teaching.

The framework’s scope is unusually broad. It addresses research writing from secondary school students in grades 9 through 12 developing their first researchable questions, through doctoral candidates building dissertations, to postdoctoral and professional researchers preparing manuscripts for peer review, across disciplines. This range matters because the temptation to offload struggle is not confined to novices. The paper opens with an anecdote about a high school student who submitted three AI-generated papers, frustrated by what the student perceived as a double standard — teachers use AI all the time, so why is it not acceptable for students? The anecdote illustrates how pervasive the temptation has become, and how the absence of clear norms leaves students improvising their own ethical boundaries. But the same dynamics apply to a seasoned researcher facing a deadline, for whom the pull toward delegation is arguably stronger because the procedural fluency to prompt effectively is already in place.

The paper’s own composition models its recommendations. Each section includes a human-AI collaboration disclaimer documenting exactly which cognitive work the researcher performed and which was delegated. The researcher conceived the paper’s focus, proposed the conceptual framework, and implemented it. Grammarly was used for editing, with sentence revision suggestions accepted or dismissed by the researcher. When Claude Pro was asked to review the initial single research question, it suggested splitting the question into two — one addressing which cognitive demands constitute extraneous load versus germane processing, the other addressing how human-in-the-loop collaboration should be designed across phases — and the researcher, after review, agreed. In building a definitions table, the researcher set the structure, identified terms, and selected sources; the AI drafted definitions grounded in those researcher-selected sources and flagged terms the researcher had overlooked. The researcher then verified each entry against the cited sources and reorganized the categories. This is the framework applied to itself: human curation and judgment retained, AI operating within boundaries the human establishes.

The transparency of these disclaimers points toward a broader implication for how AI-assisted scholarship should be evaluated. If the value of a paper lies partly in the germane processing its author performed, then disclosure of the division of cognitive labor becomes a form of methodological honesty, analogous to reporting data collection procedures. Readers can then judge for themselves whether the claims rest on understanding the author actually built or on synthesis the tool performed. The framework implies that assessment practices — in classrooms, in doctoral programs, in peer review — may need to shift from evaluating final products toward evaluating the process by which they were produced, since the product alone can no longer reveal whether the thinking occurred.

Equity considerations complicate the picture. The framework’s implications for responsive digital education include tool selection, assessment, and support for writers with cognitive differences. For some learners, certain demands that are extraneous for others may function differently, and rigid phase prescriptions could either burden or liberate them inappropriately. A student with dyslexia, for example, may experience transcription demands as genuinely extraneous friction that assistive tools should remove, while still needing to own the conceptual struggle of argument development. The framework’s emphasis on human judgment as the operational mechanism — rather than fixed rules about which tools may be used when — is what allows this flexibility. The line between offloadable and non-offloadable work is drawn by a person who understands the learner and the task, not by a policy document or a software default.

What emerges is a design philosophy rather than a prohibition. The paper does not argue that AI should be kept out of research writing; it argues that AI’s role should be calibrated to preserve the cognitive work that makes research meaningful. The productive struggle literature, drawing on work by Hiebert and Grouws and by Warshauer, positions struggle as a means of engaging in and developing critical thinking — not as an obstacle to learning but as its engine. Anne Lamott’s familiar advice to her overwhelmed brother facing an overdue report on birds — take it bird by bird — captures the framework’s practical spirit: large cognitive tasks become manageable when broken into steps, but the steps themselves must be climbed by the writer. AI can hold the ladder, hand over the field guide, and check the citations. What it cannot do is climb, because the climbing is the learning. The framework’s contribution is to specify, phase by phase, where the ladder belongs and where the writer’s feet must stay on the ground.

Subject of Research: A nine-phase human-in-the-loop framework for AI-assisted research writing based on cognitive load theory

Article Title: Designing a human-in-the-loop AI iterative process for research writing: a framework for preserving productive struggle

Article References: Chauncey, S. A. (2026). Designing a human-in-the-loop AI iterative process for research writing: a framework for preserving productive struggle. Frontiers in Computer Science, 8, Article 1947835. https://doi.org/10.3389/fcomp.2026.1947835

Image Credits: AI Generated

DOI: 10.3389/fcomp.2026.1947835

Keywords: artificial intelligence, research writing, cognitive load theory, productive struggle, human-in-the-loop, generative AI, education, academic writing, cognitive offloading, responsive pedagogy, NotebookLM, authorship

The study, carried out by Abdullah M. Alazemi of PAAET’s Language Center, used a convergent mixed methods survey design, collecting quantitative and qualitative data simultaneously through a single questionnaire administered in Arabic. The quantitative strand captured patterns of use, perceived benefits, and concerns on five-point Likert scales, while six open-ended questions invited students to describe in their own words how AI shaped their writing practices. The two datasets were analyzed separately, using descriptive statistics in SPSS and reflexive thematic analysis respectively, and then compared for convergence. The subscales showed good internal consistency, with Cronbach’s alpha values of 0.85 for AI use, 0.90 for perceived benefits, and 0.77 for concerns.

What the numbers reveal is striking in its consistency. Students reported using AI most often for language-related functions: 72.5 percent used it to correct grammar, 73.1 percent to enhance vocabulary, and 75.3 percent to express ideas more clearly in English. Meanwhile, 65.3 percent turned to AI for idea generation, 66.7 percent for summarizing academic materials, and 73.2 percent agreed that AI helped them complete writing tasks more efficiently. In other words, the technology is not primarily a shortcut for producing finished assignments; it functions across the entire arc of the writing process, from brainstorming and outlining through paraphrasing, drafting, and final revision.

The qualitative strand added texture to these patterns. Four interrelated themes emerged from the thematic analysis: AI as a facilitator of English academic writing, AI as a cognitive and academic efficiency tool, AI as a confidence-building resource, and students’ active negotiation of the opportunities and risks of AI use. Many students described a familiar frustration that AI helped resolve: knowing what they wanted to say but struggling to express it in academic English. One participant wrote that AI had helped them learn new vocabulary, formulate sentences more clearly and professionally, and increased their confidence in writing. Another described how it enabled faster organization of ideas and easier review of mistakes.

That efficiency dimension deserves particular attention. Students described using AI to condense long searches, summarize dense lectures, simplify complicated material, and organize information into manageable forms. One participant noted that AI had shortened a long journey of searching, reviewing, and writing. From a cognitive standpoint, this suggests students perceive the technology less as a ghostwriter and more as a form of cognitive support, a tool that reduces the extraneous linguistic and organizational load of academic work so that attention can shift toward understanding disciplinary content. The study interprets these practices through sociocultural learning theory, which treats tools as facilitators of participation in learning activities, positioning AI as a kind of linguistic scaffolding within students’ zone of proximal development.

Perhaps the most psychologically significant finding concerns confidence. Writing in a second language is notoriously anxiety-inducing, and 62.3 percent of participants believed AI increased their confidence in writing English, while 61.6 percent agreed it reduced writing-related anxiety. Students described having access to immediate, judgment-free feedback at any hour, something traditionally available only from teachers, peers, or writing centers during limited windows. The study draws on self-efficacy theory to interpret this pattern: as students perceived themselves becoming more capable of producing grammatically correct, coherent texts, their willingness to engage with writing tasks grew. Confidence, in these accounts, emerged not from the mere existence of AI but from a perceived reduction in the linguistic barriers that made writing feel daunting.

Yet the students themselves were far from uncritical. The fourth theme revealed a sophisticated awareness of AI’s risks. Approximately 63.4 percent of participants expressed concern about overreliance on AI tools, 58.6 percent believed AI may weaken independent writing skills, and 72.4 percent agreed that AI-generated information may sometimes be inaccurate. Some students reported deliberately reviewing, evaluating, and verifying AI outputs before using them, while others worried that habitual reliance might erode their ability to recall vocabulary or construct grammar independently. One participant cautioned that excessive dependence may reduce a student’s ability to think and solve problems independently, while another insisted that AI should serve as a supporting tool rather than a replacement for personal effort.

Notably, students intertwined ethical concerns with educational ones. Beyond worries about violating academic integrity policies, participants feared that heavy AI use would cost them opportunities to develop their own writing skills and critical thinking. The study found that 64.2 percent of participants reported concerns about the ethical use of AI in academic writing. Students effectively drew a line between responsible use, checking information, revising AI suggestions, and using AI to support understanding, and irresponsible use, copying content without verification or outsourcing thinking entirely. The issue, as they framed it, is not whether AI should be used but how.

The convergence between the quantitative and qualitative strands was strong, with no major contradictions between them. The survey data mapped the breadth of use and concern, while the open-ended responses explained the underlying experiences of language support, efficiency, and confidence. This alignment strengthens the study’s central claim: generative AI has moved beyond being an occasional educational technology and is now embedded in students’ everyday writing practices, functioning simultaneously as language tutor, cognitive assistant, and psychological buffer.

The implications reach well beyond Kuwait. The author argues that institutions should respond not with blanket prohibitions but with AI literacy frameworks that teach students to question, evaluate, and refine machine-generated content. In nursing and health sciences education specifically, guidance must address field-specific hazards, including fabricated or inaccurate clinical information and threats to patient confidentiality, and should encourage verification of any health-related content AI produces. The study’s limitations warrant caution: it was conducted at a single institution using convenience sampling and self-reported perceptions, the predominantly male sample may limit transferability, and no objective writing performance data were collected. Whether AI-assisted confidence translates into genuine long-term gains in writing competence remains an open question, one the author argues demands longitudinal research. What the study establishes firmly is that the debate over AI in higher education has shifted decisively from whether students will use these tools to how educators can ensure they use them well.

Subject of Research: Generative AI use for English academic writing among nursing and health sciences students in Kuwait

Article Title: Generative artificial intelligence use for academic writing in English among Kuwaiti nursing and health sciences students in a mixed methods study

Article References: Alazemi, A. M. (2026). Generative artificial intelligence use for academic writing in English among Kuwaiti nursing and health sciences students in a mixed methods study. Discover Education, 5(1), Article 1159. https://doi.org/10.1007/s44217-026-02297-w

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02297-w

Keywords: generative AI, academic writing, nursing education, health sciences, English medium instruction, higher education, Kuwait, ChatGPT, mixed methods, AI literacy, self-efficacy, writing anxiety

The study, published in Discover Education by Alexander M. Sidorkin, examined a single 16-week semester of a course in child and adolescent development. Thirty-six students enrolled, most of them juniors and seniors at a regional public university serving a substantial Hispanic-serving and first-generation population. Most arrived with no prior coursework in qualitative research, no familiarity with academic publishing conventions, and little experience writing extended scholarly manuscripts. Yet the course required each of them to produce a full-length qualitative research manuscript of at least 5,000 words and to submit it to a peer-reviewed journal, with confirmation of submission as part of the final assignment. This was not a simulation of professional work; it was the work itself.

The pedagogical gamble was that generative AI could absorb the procedural burden that historically made such ambitious assignments impractical for novices. Authentic assessment, the idea that students should be evaluated on tasks resembling real professional practice, has struggled for decades with a built-in tension: authentic tasks demand competencies students have not yet developed, and the scaffolding needed to bridge that gap often dilutes the authenticity it is meant to support. Sidorkin’s argument is that a capable language model changes this calculus. A student who previously could not construct a conventional methods section or navigate disciplinary terminology can now obtain that support on demand, shifting the locus of difficulty away from routine production and toward decisions about evidence, interpretation, and methodological fit.

Crucially, the course did not treat AI as a threat to be policed. Students used a custom course-specific tool, the CHDE 111 Class Companion, a configured GPT that offered navigation, concept clarification, assignment protocols, and rubric-based draft review while explicitly instructed to support learning without providing completed assignments. They were also free to use ordinary ChatGPT or other major platforms without limitation, and the syllabus placed responsibility for the quality, veracity, and originality of the final product squarely on the student. For most assignments, students submitted complete logs of their AI conversations alongside their work, making part of the human-machine interaction visible for feedback and analysis.

To measure whether unrestricted AI access compressed student outputs, the study applied three rule-based indicators to the 33 final manuscripts. The Methods Specificity Index, a 0-to-6 scale scoring the presence of six methodological components such as a named qualitative approach, a concretely specified data source, sampling logic, analytic procedure, ethical safeguards, and a rationale connecting method to question, showed dramatic upper-bound compression. The median score was the maximum of 6, nearly 70 percent of manuscripts received the full score, and 97 percent scored either 5 or 6. Methodological specification, in other words, had become something AI scaffolding could reliably deliver.

The other two indicators told a different story. Claim-Evidence Coupling, the proportion of interpretive statements locally accompanied by specific evidentiary support such as a quotation, citation, or reference to the study data, ranged from 0.000 to 0.750 with a median of 0.333. Theory-Interpretation Linkage, the proportion of interpretive statements that explicitly invoked the manuscript’s own theoretical framework, ranged from 0.000 to 0.917 with a median of 0.444. Both remained substantially dispersed across the corpus. Students differed enormously in whether their interpretive claims were tethered to actual evidence and whether theory genuinely informed their analysis, dimensions that AI can assist with but cannot resolve, because warranting a claim depends on the actual data and argument of each individual project.

This differential pattern is the study’s central empirical finding. Compression appeared where structural scaffolding suffices; dispersion persisted where situated judgment is required. The theoretical framework behind the study, drawing on the five-dimensional authentic assessment model of Gulikers and colleagues, holds that when professionals routinely use AI, an assessment that categorically excludes it may actually resemble professional work less faithfully. Student contribution, in this view, is not independent text production but the direction, evaluation, integration, and revision of work produced within a human-AI system, a capacity the study terms extended executive cognition, paired with discerning thinking, the ability to judge AI output for substance and fit rather than equating fluency with quality.

Detailed comparison of four contrasting students, two from the top and two from the bottom of the course-points distribution, illuminated what those differences look like in practice. The stronger students orchestrated their AI use: they specified tasks precisely, supplied information the model could not infer, such as which social media discourse threads belonged in the analysis, repeatedly asked the Companion to audit sections for coherence, and pushed through multi-turn revision cycles until the conceptual fit was right. One student rejected a Companion-generated framework summary as misaligned with her research question and iterated across several turns until satisfied. The weaker students, by contrast, engaged in sparse, one-directional interactions, accepting template language and structural outlines without constraining them to the specifics of their actual studies. Their manuscripts could look competent at a glance, but theme statements were thin on evidence and theory was named rather than deployed.

The study is candid about its limits. It is a single-semester case study without a comparison group, so it cannot isolate AI orchestration from prior preparation, effort, or research aptitude. The three indicators are heuristic measures applied with the assistance of ChatGPT rather than validated instruments, and the instructor, course designer, tool designer, analyst, and author are all the same person, a concentration of roles that creates real confirmation-bias risk despite safeguards such as fixed rules and independent case selection. The process evidence also covers only a late slice of the semester, so it cannot show whether orchestration practices developed through instruction or simply differed among students from the start.

Even with those caveats, the implications are significant for anyone designing courses in the AI era. The study suggests that authentic assessment can remain viable and diagnostically useful when instructors explicitly distinguish between dimensions AI can scaffold reliably and dimensions where student judgment remains consequential, then align tasks, feedback, and rubrics accordingly. It also suggests that AI orchestration itself should become an instructional object, modeled and taught rather than assumed, and that when AI can generate prose quickly, the bottleneck shifts toward reading, evaluating, and revising it. Whether the pattern transfers across disciplines and institutions remains an open empirical question, but the Sacramento case demonstrates that giving every student a powerful AI assistant did not make them interchangeable. The tool was common; the judgment was not.

Subject of Research: Authentic assessment and student judgment in an AI-integrated undergraduate qualitative research methods course

Article Title: Authentic assessment in an AI-integrated qualitative research methods course

Article References: Sidorkin, A. M. (2026). Authentic assessment in an AI-integrated qualitative research methods course. Discover Education, 5(1), Article 1153. https://doi.org/10.1007/s44217-026-02273-4

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02273-4

Keywords: authentic assessment, generative AI, qualitative research methods, undergraduate education, evaluative judgment, human-AI collaboration, assessment design, rubric, extended executive cognition, pedagogy, case study, academic writing

The study, published in the journal Discover Education, was led by Muhammad Takwin Machmud of the Department of Community Education, together with colleagues Nurhudayah Manjani and Natalia Silalahi. The team set out to answer a deceptively simple set of questions: how well do university students understand plagiarism, how often do they engage in plagiarism-related behaviors when completing assignments, how do they use plagiarism detection tools, and how do their similarity scores compare with the benchmarks their university imposes?

When asked to define plagiarism in their own words, 81 of the 105 surveyed students, or 78.6 percent, produced answers that matched the expected conceptual keywords, such as stealing or duplicating other people’s work, failing to credit the original author, or claiming someone else’s work as one’s own. On the surface, that sounds encouraging. But the researchers noted that most students framed plagiarism narrowly, as literal copy-pasting, while only a small minority described it in broader terms such as academic dishonesty or violations of intellectual ethics. The team also acknowledged a caveat: because the questionnaire was administered without supervision, some students may have looked up definitions online before answering, a limitation consistent with prior research showing that self-reported measures of academic integrity knowledge can be inflated by external assistance or socially desirable responses.

The behavioral data told a more complicated story. A majority of students, 61.9 percent, reported that they only occasionally paraphrase when adapting information from original sources, while 31.4 percent said they always paraphrase and just 6.7 percent said they never do. Direct copying was far more common than the researchers anticipated: 87.6 percent of students admitted to occasionally copying information directly from sources, and 5.7 percent said they always do. When students were interviewed about why, many explained that copying had become a routine time-management strategy, even though they still cited the original authors. The researchers interpreted this as evidence that many students simply struggle to paraphrase properly, hampered by heavy assignment loads, difficulty understanding dense reading materials, limited vocabulary, and, by their own admission, occasional laziness.

Citation practices, by contrast, appeared relatively healthy. More than half of the surveyed students, 58.1 percent, said they always credit the original author, 37.1 percent said they do so occasionally, and only 4.8 percent reported never citing sources. In other words, students are not wholesale thieves of ideas; they are writers who attribute their sources but often fail to transform them, reproducing sentence structures and phrasing so close to the original that detection software flags them anyway.

To move beyond self-reports, the researchers analyzed 30 scientific article assignments written for a Research Methodology course, running each document through university-approved plagiarism detection software under standard institutional settings that excluded bibliographies and quoted text and ignored small matches below one percent. The results were striking. Similarity index scores ranged from 5 percent to 85 percent, with an average of 41 percent and a median of 36.5 percent, and a standard deviation of roughly 22.5 percent indicating enormous variability in writing practices. Against the university’s benchmark, which caps similarity at 35 percent for undergraduate students in the social sciences and humanities, only 14 of the 30 assignments passed. Sixteen, or 53.3 percent, exceeded the threshold.

Perhaps the most technically revealing finding concerned where the matched text came from. Internet sources accounted for the largest share of similarity matches, averaging 39.7 percent, followed by previously submitted student papers at 18 percent and academic publications at just 12.4 percent. This pattern suggests that students lean heavily on open web resources when completing assignments, and that some flagged text may echo work already sitting in detection databases from earlier student submissions. The researchers connect this to broader concerns raised in the literature: the sheer abundance of online material and the ease of accessing it have been cited as key drivers of rising plagiarism, and reports from Indonesia indicate that student plagiarism cases climbed by nearly 70 percent in 2017 alone.

The study also examined how students actually use plagiarism detection tools, and here the findings were more optimistic. Turnitin dominated, used by 73.3 percent of surveyed students, well ahead of Grammarly at 36.2 percent, Duplichecker at 21 percent, Unichecker at 11.4 percent, and other tools at 8.8 percent. Students gravitated toward paid software, explaining that the university provided premium access and that paid platforms offer richer features: vast databases spanning journals, books, magazines, and webpages; detailed similarity reports identifying the original sources of matched text; highlighted passages of concern; and support for many document formats. Turnitin’s own database, according to figures cited in the study, contains 1.9 million student papers, 190 million subscription articles, content from 1,500 publishers worldwide, and coverage of 97 percent of the top 10,000 journals.

Crucially, the researchers found that detection tools function not merely as gatekeepers but as revision aids. Analyzing institutional records from 24 final-year social science students who defended their theses between January and June 2025, the team documented students running multiple checks to drive their scores down. One student cut a similarity index from 42 percent to 13 percent between the first and second checks; another dropped from 40 percent to 14 percent; a third ran three checks, gradually descending from 28 percent to 26 percent to 25 percent. Across all 24 theses, final similarity indices ranged from 5 percent to 31 percent, averaging 22.8 percent, a level most students achieved only after iterative revision. The software, in effect, had become part of the writing process itself, with 82.9 percent of surveyed students using it for scientific articles, 52.4 percent for academic reports, and 39 percent for daily assignments.

The authors caution that a similarity index is an indicator of textual overlap, not proof of misconduct, since properly cited quotations, references, and common academic phrases can inflate scores, and that thresholds vary widely across institutions and publishers, with no single universal standard. They also note challenges such as false positives and privacy concerns that universities must manage through clear policies and training. Still, their conclusion is pointed: plagiarism among university students should be treated not only as an ethical failing but as an academic literacy problem demanding pedagogical solutions, from explicit instruction in paraphrasing, quoting, and summarizing to formative checking systems that let students revise before final submission. At Universitas Negeri Medan, that philosophy is already institutionalized, with a dedicated Plagiarism Detection Team, tiered similarity thresholds by degree level and discipline, mandatory originality statements, and library-run programs such as Kelas Literasi, a class devoted to avoiding plagiarism through better citation and writing practice. Detection software, the study suggests, raises awareness and saves instructors time, but lasting change will come from teaching students how to write with sources, not merely how to avoid being caught.

Subject of Research: University student plagiarism behavior and the use of plagiarism detection tools in academic writing

Article Title: University student plagiarism behavior and the role of plagiarism detection tools

Article References: Machmud, M. T., Manjani, N., & Silalahi, N. (2026). University student plagiarism behavior and the role of plagiarism detection tools. Discover Education, 5(1), Article 1004. https://doi.org/10.1007/s44217-026-02180-8

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02180-8

Keywords: plagiarism, academic integrity, university students, plagiarism detection software, Turnitin, similarity index, paraphrasing, citation practices, higher education, academic writing, Indonesia, educational research

The study was conducted between August and October 2025 using an electronic survey distributed through open online channels, which means an individual-level response rate could not be calculated, a limitation the authors acknowledge. Still, the breadth of institutional coverage gives the data unusual weight for a convenience-sampled survey. Students from both public and private universities took part, and the researchers found no meaningful gap between the two sectors in how many students had at least considered research: 80.0 percent of public university students versus 84.8 percent of private university students had thought about engaging in scientific research, a difference that was not statistically significant. Interest, in other words, is not the problem. The pipeline between interest and action is.

That pipeline was examined with considerable methodological care. Because existing validated questionnaires were not suitable for the Turkish undergraduate context, the study team built a study-specific instrument from scratch. It was developed by medical students who are themselves actively involved in research, informed by the published literature and by their direct experience of overcoming common research-related difficulties, and adapted to the realities of Turkish medical education. Item content was reviewed by a research-active professor to assess relevance and coverage, and face validity and feasibility were tested in a pilot with 20 medical students who evaluated clarity, comprehensibility, and completion time. Incomplete or inconsistent responses were excluded before analysis.

The statistical approach combined descriptive statistics with Mann–Whitney U tests for continuous variables and Chi-square or Fisher’s exact tests for categorical variables. To isolate the factors that genuinely predict research intention, the team used multivariable binary logistic regression, a technique that estimates the independent effect of each variable while holding the others constant. Statistical significance was set at p < 0.05. The results of that modeling are the heart of the study. Higher English proficiency was associated with roughly three times the odds of intending to participate in scientific research, with an adjusted odds ratio of 3.07 and a 95 percent confidence interval of 1.96 to 4.83. The presence of a research-active person in the student’s environment nearly doubled the odds, at an adjusted odds ratio of 1.92 (95 percent CI 1.30 to 2.82). Prior research experience raised the odds by a factor of 1.83 (95 percent CI 1.24 to 2.69), and prior academic-writing training by a factor of 1.86 (95 percent CI 1.10 to 3.16).

Each of these four factors tells a different story about what is missing. English proficiency is perhaps the least surprising but most consequential. The overwhelming majority of biomedical literature is published in English, and the dominant reference databases, journals, and international conferences operate in that language. A student who cannot comfortably read an English-language paper, let alone write one, faces a structural barrier that no amount of enthusiasm can fully compensate. The finding that English proficiency roughly triples the odds of research intention suggests that language instruction is not merely a general educational nicety but a direct gateway into the scientific enterprise itself.

The environmental factor is the most sociologically interesting. Simply knowing or being near someone who is actively doing research nearly doubles a student’s odds of intending to do research themselves. This is consistent with long-standing observations in the sociology of science that research careers are often seeded by proximity: a classmate who publishes, a teaching assistant who runs a study, a physician who mentions a case report over coffee. Such contacts demystify the process, provide informal mentorship, and signal that research is something people like the student actually do. In a large, geographically dispersed medical education system like Türkiye’s, the uneven distribution of such role models may help explain why research engagement remains limited overall despite broad interest.

The publication data in the study quantify how narrow that pipeline really is. Among all 1,062 respondents, only 10.8 percent reported at least one peer-reviewed publication. Among the subset who had actually participated in research, 23.7 percent had published. That means roughly three out of four students who managed to get involved in a research project still never saw the work reach print. The authors treat prior research engagement and publication history as separate indicators of previous research exposure and research output, and the gap between the two is itself a finding. Getting students into research teams is only half the battle; converting participation into a finished, peer-reviewed product requires skills, time, and supervision that many students evidently never receive.

The secondary objectives of the study add another layer. The researchers also examined factors associated with knowledge of reference-management software, the tools such as citation managers that are basic infrastructure for anyone writing a manuscript. In an era when artificial intelligence is reshaping how literature is searched and summarized, the ability to organize and cite sources systematically remains a foundational skill, and the fact that the authors singled it out for analysis signals how unevenly distributed even basic research tooling is among undergraduates. The study-specific questionnaire was designed to capture these research-related knowledge and perception gaps that existing instruments, built for other contexts, had left underexplored.

The authors are careful about what their design can and cannot show. Because the survey is cross-sectional, capturing a single moment in time, the findings should be interpreted as associations rather than causal relationships. A student with high English proficiency may intend to do research, or students who decide to pursue research may then invest in their English; the data cannot distinguish these directions. Convenience sampling through open online channels likewise means the sample may overrepresent students who are already more connected and more research-curious. The authors received no financial support for the study, obtained ethical approval from the İstinye University Ethics Committee under approval number 2025-330, conducted the work in accordance with the Declaration of Helsinki, and collected all data anonymously with electronic informed consent from every participant.

Even with those caveats, the practical implications are hard to ignore. If intention to do research is driven by English skills, writing training, hands-on experience, and contact with research-active people, then medical schools have four concrete levers to pull. Embedding academic-writing instruction into the undergraduate curriculum, creating structured early research experiences, expanding English for medical purposes courses, and deliberately connecting students with mentors and peers who are actively publishing could each move the needle. The study’s own authorship illustrates the point: the questionnaire was built by students who had themselves navigated these barriers, and the paper’s author list spans institutions across Türkiye and beyond. The message from the data is that research careers rarely ignite in isolation. They catch fire where language, mentorship, training, and opportunity happen to collide, and the task for educators is to make that collision far less accidental.

Subject of Research: Barriers to research engagement among medical students in Türkiye

Article Title: Barriers to research engagement among medical students in Türkiye: the role of environmental factors

Article References: Kabakcı, A. İ., Yılmaz, B. N., Şahin, B., Canpolat, Y., Deiri, M. A., Demirbağ, K., Kaba, İ., Kamar, Z., Kuzay, R., Şimşek, E. C., Tozcu, M. D., Uyanık, F. B., & Güngör, A. (2026). Barriers to research engagement among medical students in Türkiye: the role of environmental factors. BMC Medical Education. https://doi.org/10.1186/s12909-026-10514-1

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10514-1

Keywords: medical students, research engagement, medical education, Türkiye, mentorship, English proficiency, academic writing, cross-sectional study, environmental factors, publication output, survey research, logistic regression

Kim, writing from Atlanta at the age of seventy-six and drawing on a formation that spans Korea and America, Christian theology and Chinese thought, describes working with AI as being surrounded by thirty exceptionally capable doctoral students. One offers a term, another restructures the argument, a third suggests what the writer really means, and others return carrying notes from Peirce, Wittgenstein, Confucius, or contemporary sociolinguistics. The experience is exhilarating, and Kim does not deny it. But the metaphor breaks down at a crucial point. Thirty real doctoral students are thirty centers of resistance. One will dislike the argument, another will say Peirce has been misunderstood, a third will accuse the writer of romanticizing East Asia. Someone, sooner or later, will deliver the sentence every scholar needs to hear: the work is beautiful, and it is also wrong.

AI can imitate such objections, but imitation of plurality is not plurality. Current systems, Kim argues, inhabit overlapping textual worlds and share habits of exposition. They are gifted at coherence. Give them two distant ideas and they will often find a bridge before asking whether one ought to exist. This, in the essay’s framing, may be generative AI’s characteristic epistemic danger: the texture of consensus without the evidential weight of consensus. Five machines admiring an idea do not constitute peer review, yet five machines can help that idea put on a suit and tie by afternoon. The old difficulty of scholarship was that a living question might die before a scholar could pursue it. The new difficulty is that a foolish question can become a polished manuscript before dinner.

The essay’s sharpest formulation concerns what AI has actually democratized. Public discussion often claims that these tools have democratized intelligence. Kim counters that they have done something more unsettling: subsidized the appearance of intelligence. Eloquence can now be rented. Erudition can be rented. Structure can be rented. Even the external appearance of judgment can be rented by the hour, or by the token. This does not mean the renter is stupid, and Kim is emphatic on that point, describing a daily practice of using AI gratefully as an extraordinary workshop. A question that once required weeks of preparation can now be tested while it is still warm. Confucian zhengming, the rectification of misalignment between name and reality, can be placed beside Peircean semiotics, Wittgenstein can be invited to disturb the arrangement, and a sociolinguist’s objection can be summoned in seconds. The library is illuminated at three in the morning. The translator is awake.

Speed, Kim acknowledges, preserves the temperature of a question, and that is no small gift. But speed also creates a new temptation: linguistic completion before intellectual completion. The machine closes sentences, fills conceptual gaps, smooths awkward transitions, and reconciles tensions with extraordinary politeness. What once remained visibly unresolved can now be made to look settled. Fluent closure has become almost free, and the institutional consequences follow directly. When half-formed thoughts can be dressed for submission in an afternoon, journals inherit part of the cost that AI has removed from authors. Editors and peer reviewers become unpaid inspectors of premature coherence. The manuscript becomes cheaper to produce precisely when judgment becomes more expensive to supply.

Kim’s personal response is to cultivate a virtue that sounds almost trivial but carries considerable technical weight: the ability to say almost. Almost right. Almost clear. Almost mine. Almost worthy of publication. The discipline involves a battery of resistant responses to fluent output: the draft is too smooth, the problem has been Westernized, the concepts should not be reconciled yet, the word is technically correct but existentially wrong, the English has been improved while the thought has been weakened, the concept is right but the center of gravity is wrong. Kim increasingly suspects that this capacity will matter more than brilliant prompting. Prompt engineering, in the essay’s wry framing, is becoming a minor priesthood, complete with incantations and claims to esoteric skill. Learn the spells by all means, but the deeper human competence is recognizing when an elegant answer has betrayed the question.

The problem is especially acute for thinkers who work across languages and civilizations. A thought may arrive with a Korean emotional temperature, a Chinese conceptual skeleton, a Christian theological horizon, and a Western philosophical question, and academic English asks it to pass through customs. AI is a magnificent customs broker, Kim writes. It can move the cargo quickly. It can also quietly replace the cargo. To guard against that substitution, Kim has adopted a deliberately primitive discipline: before inviting AI into an important thought, two or three ugly sentences of the original intuition are preserved on the page, sometimes in Korean, sometimes grammatically incomplete. The purpose is not legal ownership, since the mythology of the solitary author was always exaggerated and people think with teachers, enemies, books, languages, lovers, traditions, and ghosts. The purpose is intellectual provenance, the ability to trace a refined phrase back to the crude perception from which it grew.

That provenance question is illustrated with a phrase from Kim’s own writing, indexicality without occupancy, and the Korean intuition behind it: AI speaks like a person of authority, but there is no one behind the words to bear responsibility; there is a sign, but no person there. The important question, Kim argues, is not whether the machine offered the elegant phrase or whether the human refined it through forty exchanges, but whether the language clarified the original intuition or replaced it. Formal governance will not settle this. Academia will need disclosure rules for AI use, publishers will invent policies, and universities will produce forms, all necessary and all predictably bureaucratic. The more intimate problem will remain unregulated: whether scholars will still recognize their own thoughts after the machines have improved them.

Kim is less worried that AI will become intelligent than that human beings will become satisfied with the signs of intelligence, and academia already rewards those signs: a polished abstract, correct vocabulary, a symmetrical framework, six boxes in a diagram, a confident conclusion beginning with a call for fundamental reconceptualization. AI did not invent academic vanity, the essay observes; it simply industrialized some of its preferred costumes. The antidote is not nostalgia, and Kim explicitly refuses to dismiss a technology that enables what the essay calls speculative adjacency, bringing distant bodies of thought into rapid, exploratory contact. When the hallways become short enough, the architecture of thinking changes. But the cheaper inquiry becomes, the more expensive discernment must become. Not every connection deserves development, not every curiosity deserves a paper, and not every publishable paragraph contains a thought worth publishing. In the old economy of scholarship, scarcity imposed crude discipline through limited books, research trips, conversations, and years. AI has removed many of those restraints, and what is needed now, Kim writes, is intellectual fasting.

The essay closes on a question that reframes the entire debate. Rather than asking whether AI will make us more intelligent, a question that flatters machine and human alike, Kim asks whether, surrounded by nearly free fluency, we are becoming more discriminating. Are our questions improving? Can we recognize a false harmony? Can we preserve an inelegant intuition long enough to discover what it means? Can we throw a curveball that breaks the window of premature coherence and then inspect the glass? Above all, can we still say almost when the machine gives us exactly the beautiful answer we hoped to hear? That small word, Kim suggests, may become one of the last defenses of formed intellect. AI has made fluency cheap. Judgment is now expensive, and we should spend accordingly.

Subject of Research: The epistemic impact of generative AI on human judgment, intellectual provenance, and scholarly writing

Article Title: The ability to say “almost”

Article References: Kim, S. (2026). The ability to say “almost”. AI &amp; SOCIETY. https://doi.org/10.1007/s00146-026-03297-y

Image Credits: AI Generated

DOI: 10.1007/s00146-026-03297-y

Keywords: generative AI, artificial intelligence, epistemology, academic writing, peer review, judgment, prompt engineering, intellectual provenance, AI & Society, scholarship, language models, critical thinking

The concern emerges against a backdrop of rapid institutional response to generative AI. As large language models have become widely available, many journals, universities, and professional bodies have adopted policies that limit or restrict the use of AI in academic writing. Enforcement of these policies has leaned heavily on AI detection software, tools that attempt to distinguish machine-generated prose from human writing by analyzing stylistic patterns, vocabulary choices, and overall cadence. The problem, as Smith and Aggarwal emphasize, is that these tools are far from infallible. Citing recent research, they note that even automated detectors do not achieve perfect accuracy, and that human readers perform considerably worse. One study found that people cannot identify AI-generated writing any more reliably than random chance would allow.

Despite these limitations, anxiety about being accused of using AI is demonstrably shaping how authors write. The em dash has been singled out in public discourse, including commentary in major newspapers, as a marker of perceived overuse by AI tools. Yet the authors point out that peer-reviewed studies establishing a genuine association between the punctuation mark and machine generation are lacking. The association is largely a matter of perception, amplified by social media conversation and anecdote rather than systematic evidence. Still, perception can be powerful. Because the em dash may now carry a negative connotation, authors who wish to avoid evoking skepticism from reviewers and editors may simply stop using it.

The pressure is not hypothetical. In preparing a prior manuscript, one of the authors was advised by a mentor that the use of em dashes might be interpreted by reviewers as a sign of AI involvement. Both authors report hearing similar accounts from colleagues across academic medicine. These anecdotes suggest that a quiet recalibration of writing style is already underway, driven not by journal policy or formal guidance but by fear of suspicion. The authors argue that trainees and early-career academics, who are still developing their scholarly voices, may be particularly vulnerable to these pressures, internalizing the idea that certain stylistic choices are risky before they have had the chance to make those choices their own.

What is at stake, the correspondence contends, extends well beyond a preference for one punctuation mark over another. Writing, like other forms of art, is a medium through which individuals express personality, share perspective, and communicate thought. Stylistic signatures, including rhythm, cadence, and punctuation habits, are part of what makes an author’s voice recognizable. By unnecessarily restricting these elements, the academic community risks constricting what scholars are able to say and how they are able to connect with readers. The authors suggest this may be particularly damaging in fields such as psychiatry, where narrative skill and clarity of expression are not decorative extras but essential professional tools, central to both clinical communication and scholarship.

The technical reality of AI detection undercuts the logic of policing individual stylistic features. As Smith and Aggarwal observe, even automated detection software does not rely on any single tell in isolation. Instead, such systems perform a holistic analysis of a piece of writing, weighing overall style, paragraph length, grammatical choices, and vocabulary together before rendering a judgment. If the most sophisticated detectors require whole-text analysis to reach conclusions of imperfect reliability, then treating a lone punctuation mark as evidence of machine authorship is analytically indefensible. Overemphasis on a single element, such as punctuation considered in isolation, raises the very real potential for false accusations against human authors whose only offense is a fondness for the em dash.

The authors’ proposed alternative is straightforward: rather than asking authors to change how they write, the field should emphasize and demand transparency about AI use. This means clear authorship guidelines and explicit journal policies governing the use of generative AI, together with education for scholars on the importance of disclosure. Transparency, in this framing, addresses the actual problem, which is undisclosed machine assistance, rather than punishing stylistic expression that happens to overlap with the output of language models. The approach aligns with emerging guidance in medical education publishing, including recent international guidance on when and how to disclose AI use in academic publishing.

At the same time, the correspondence does not dismiss the need for safeguards. Smith and Aggarwal acknowledge that protections surrounding AI use are vitally necessary to prevent the field from becoming over-reliant on technology, with the risks to originality, accountability, and scientific integrity that such reliance would entail. Their argument is one of proportion: safeguards should target disclosure and integrity, not the erosion of individual voice. The distinction matters because style-based policing is both ineffective, given the weakness of the evidence linking any single feature to AI, and corrosive, given that it teaches a generation of writers to flatten their prose into something inoffensive and anonymous.

The em dash controversy is a small episode with outsized symbolic weight. It captures a moment in which institutions, reviewers, and writers themselves are groping toward norms for a technology that arrived faster than the rules governing it. The message from these psychiatry educators is a caution against premature judgment: with imperfect detection software, even less accurate human judgment, and only a weak and largely anecdotal association between a punctuation mark and machine text, the field should not allow a perceived association to change how people write. How we communicate, they argue, is too important to be dictated by suspicion of a dash.

Subject of Research: The impact of perceived AI-associated writing styles on academic authorship and publishing

Article Title: The Em Dash—Should We Allow AI to Change Our Writing Styles?

Article References: Smith, A. C., & Aggarwal, R. (2026). The Em Dash—Should We Allow AI to Change Our Writing Styles?. Academic Psychiatry. https://doi.org/10.1007/s40596-026-02435-4

Image Credits: AI Generated

DOI: 10.1007/s40596-026-02435-4

Keywords: em dash, artificial intelligence, academic writing, AI detection, writing style, academic publishing, authorship, transparency, psychiatry education, generative AI, scholarly voice, peer review

The study, conducted by Muhammad Shaban Rafi of Riphah International University and Lubna Khalil of the University of the Punjab, applied the Artificial Intelligence Assessment Scale developed by Perkins and colleagues in 2024 to evaluate the appropriate role of generative AI across the successive stages of dissertation writing. Rather than treating AI as either a threat or a miracle, the framework allowed the researchers to distinguish between tasks that can be fully delegated to machines, tasks that require a human–AI partnership, and tasks that must remain exclusively human. The results challenge both the alarmist narrative that AI will hollow out doctoral scholarship and the utopian claim that it can write a dissertation on command.

Where the technology shone most clearly was in the mechanics of writing. Participants reported that generative AI could be used without reservation to improve grammar, sentence structure, and overall coherence, enhancing the clarity and efficiency of their prose. For many doctoral students, particularly those writing in a second language, this mechanical support removed a persistent barrier that had previously consumed hours of revision. By offloading surface-level editing to the machine, students said they could redirect their energy toward deeper intellectual engagement with their arguments, literature, and data, a shift that supervisors in the study viewed as a legitimate pedagogical gain rather than a shortcut.

The second major application involved data analysis. Doctoral candidates described using generative AI to help analyze large qualitative datasets by defining a coding frame, identifying recurring trends, and conducting sentiment analyses. In the social sciences and humanities, where interviews, open-ended survey responses, and archival texts can generate overwhelming volumes of material, the ability of large language models to propose initial coding categories and flag patterns offered a practical starting point. Crucially, the participants did not describe the AI as replacing their analytical judgment. Instead, the machine-generated coding frames served as provisional scaffolding that researchers then tested, refined, and validated against their own close reading of the data.

A third productive territory was argument structuring. Students reported using generative AI to organize literature reviews, suggest logical arrangements of sentences and paragraphs, and generate counterarguments that stress-tested their claims. The capacity of these models to simulate an opposing viewpoint proved especially valuable in the humanities, where a dissertation’s strength often depends on anticipating objections. By prompting the system to challenge a thesis statement or identify gaps in a literature synthesis, candidates effectively gained a tireless sparring partner available at any hour, one that could surface alternative framings the writer had not considered.

Yet the enthusiasm had sharp limits. Participants were emphatic that generative AI should be limited or prohibited in areas demanding critical reasoning, originality, and cultural context. The core intellectual contributions of a dissertation, formulating research questions, interpreting findings within their disciplinary traditions, and producing genuinely novel insights, were judged to require human intelligence at their center. Supervisors in particular warned that AI-generated content may lack accuracy and contextual depth, producing text that reads fluently but fails under scholarly scrutiny. The study highlights the well-documented problem of hallucination in natural language generation, in which models confidently produce plausible-sounding but false claims, a hazard especially dangerous in academic contexts where fabricated citations or mischaracterized sources can constitute serious misconduct.

This is where the concept of prompt literacy enters the picture. The researchers argue that the quality of output from generative AI is not a fixed property of the technology but a function of how skillfully the user frames the request. Prompt engineering has emerged in recent years as a recognized digital competence, with studies describing it as a new twenty-first-century skill and cataloging systematic techniques, from chain-of-thought prompting, which guides models through explicit reasoning steps, to retrieval-augmented generation, which grounds model outputs in external source documents. The study positions prompt literacy as an essential component of broader AI literacy, encompassing the ability to specify role, context, constraints, and desired output format in a prompt, and to iteratively refine instructions when the first response falls short.

The pedagogical implications are significant. If the difference between a vague, hallucination-prone response and a precise, useful one lies in the prompt, then universities face a choice about whether to teach these skills explicitly or leave students to learn by trial and error. The study responds by providing a scale for the appropriate use of generative AI at each stage of doctoral writing, along with ready-to-use prompts developed as part of the research. Such resources suggest a model of AI integration in which institutions define clear boundaries, students develop the technical and critical competence to work within them, and assessment practices evolve to distinguish between acceptable assistance and academic dishonesty. The ethical dimension is central: the study received ethical approval under professional research ethics codes, and its framework is explicitly designed for the ethical integration of AI in educational assessment.

The findings arrive amid a rapidly expanding body of research on generative AI in higher education, including studies of students’ prompt patterns, hermeneutic approaches to prompt optimization, and surveys of hallucination in language models. What distinguishes this study is its focus on the doctoral dissertation, the genre in which the stakes of originality are highest and the tension between efficiency and authenticity most acute. By grounding its conclusions in the lived experience of students and supervisors rather than speculation, it offers a pragmatic middle path: generative AI as a powerful tool for mechanics, data handling, and argument testing, governed by human judgment, validated at every step, and wielded by researchers who understand both what to ask and what the machine cannot be trusted to answer.

For the current generation of doctoral candidates, the message is double-edged. The technology can genuinely save time and elevate the quality of dissertation writing, but only for those who invest in the literacy to use it well. As generative AI becomes embedded in the infrastructure of research, the study suggests that the most important qualification for the twenty-first-century scholar may not be the ability to write without AI, or even with it, but the ability to converse with it critically, skeptically, and skillfully, keeping human intelligence, as the authors insist, at the core of the scholarly enterprise.

Subject of Research: The applications, limitations, and prompt literacy requirements of generative AI in doctoral dissertation writing

Article Title: Generative AI in Doctoral Dissertation Writing: Applications, Limitations, and the Need for Prompt Literacy

Article References: Rafi, M. S., & Khalil, L. (2026). Generative AI in Doctoral Dissertation Writing: Applications, Limitations, and the Need for Prompt Literacy. Frontiers of Digital Education, 3(2), Article 14. https://doi.org/10.1007/s44366-026-0088-9

Image Credits: AI Generated

DOI: 10.1007/s44366-026-0088-9

Keywords: generative AI, doctoral dissertation writing, prompt literacy, AI Assessment Scale, human–AI collaboration, AI literacy, higher education, academic writing, prompt engineering, hallucination, ethical use of AI, social sciences and humanities

The collaborative intelligence perspective that anchors the study draws on a body of theory in human-computer interaction and learning science which holds that the most productive human-AI partnerships are those in which the human retains epistemic authority while the machine supplies speed, breadth, and fluency. In writing tasks, this means the AI may suggest phrasing, generate outlines, explain grammar, or summarize sources, but the student must decide what counts as an argument, which claims are defensible, and how the text should be organized for a specific audience. When this division of labor works well, the student learns more efficiently because cognitive load is reduced on low-level mechanical tasks and redirected toward higher-order thinking. When it works poorly, the student risks accepting fluent but shallow or inaccurate output, and the writing becomes a patchwork of machine text that the author does not fully understand or endorse. The Indonesian context adds a distinctive layer to this dynamic, because most generative AI systems are trained predominantly on English-language data while Indonesian students often write in English as an additional language.

English-medium academic writing is a requirement in many Indonesian universities, particularly for students preparing journal articles, theses, and international conference papers. For these students, generative AI presents an alluring proposition: a tireless, instant, and free editor that can smooth grammar, suggest academic vocabulary, and restructure sentences. The study situates itself in this reality, recognizing that Indonesian students arrive at AI tools with genuine practical needs. At the same time, universities in Indonesia, as elsewhere, have scrambled to formulate policies on AI use, with some institutions prohibiting it outright and others encouraging disciplined experimentation. The absence of settled policy creates an environment in which students largely regulate their own AI use, guided by peer norms, instructor expectations, and their own judgments about academic integrity. Understanding what students actually do, rather than what policies assume they do, is therefore an essential empirical task, and it is precisely the gap this research sets out to fill.

Methodologically, the study adopts an assessment-oriented design, gathering evidence of how students interact with generative AI during authentic academic writing activities. Rather than relying solely on surveys about attitudes, the researchers examine the texture of the interaction itself: how students prompt the AI, how they evaluate its suggestions, when they accept, reject, or modify output, and how the final written products reflect the negotiation between human intent and machine generation. This interactional focus is what distinguishes a collaborative intelligence analysis from a simple usage study. The unit of analysis is not the tool and not the student in isolation, but the dyad, the evolving partnership through which a text comes into being. Assessment in this framework is not merely grading the final essay; it is diagnosing the quality of the collaboration, including the student’s critical engagement, revision behavior, and willingness to exercise authorship over machine-generated material.

The findings, as framed by the researchers, point toward a picture of students as active negotiators rather than passive consumers of AI text. Students in the study engaged with generative AI for a range of sub-tasks within writing, from brainstorming topics and generating thesis statements to checking grammar and rephrasing awkward passages. Crucially, the collaborative intelligence lens reveals that the value of these interactions depends heavily on the student’s own disciplinary knowledge and language confidence. Students with stronger command of their subject and of academic English were better positioned to interrogate AI output, spot errors, and reshape suggestions to fit their arguments. Students with weaker foundations were more likely to accept output uncritically, which the study treats as a pedagogical warning sign. This pattern reinforces a conclusion that is emerging across the international research literature: AI literacy cannot be separated from domain literacy, and tools that promise to compensate for weak knowledge may instead entrench it if critical evaluation is absent.

One of the most technically interesting dimensions of the study is its attention to prompting behavior, the set of instructions students type into generative AI systems to elicit output. Prompting is, in effect, a new form of academic skill, blending elements of information retrieval, task specification, and iterative dialogue. Effective prompters provide context, specify audience and register, request alternatives, and feed back corrections across multiple turns. Novice prompters issue vague commands and receive generic text. The study’s interactional data show that Indonesian students vary widely along this spectrum, and that prompting quality is itself learnable. This observation has direct implications for curriculum design. If prompting, evaluating, and integrating AI output are teachable skills, then universities can respond to generative AI not with bans but with instruction, embedding AI interaction training inside existing writing courses where students can practice under guidance and receive feedback on both their prose and their process.

The collaborative intelligence perspective also reframes the perennial anxiety about academic integrity. From this viewpoint, the central question is not whether a student used AI but whether the student can account for the text, defend its claims, and explain its reasoning. A student who uses AI to polish grammar while retaining full ownership of ideas and structure is engaged in a fundamentally different activity from one who submits generated text on a topic they have never researched. The study argues that assessment practices should be designed to make this distinction visible, through oral defenses, process portfolios, draft histories, and in-class writing components that reveal the student’s actual competence. Such approaches align assessment with the reality of professional writing in the AI era, where working academics and professionals already use these tools under norms of disclosure and accountability.

The broader significance of the research extends beyond Indonesia. The country hosts one of the largest higher education systems in Southeast Asia, with tens of millions of students, and its experience with generative AI in writing offers lessons for any education system serving multilingual learners. Generative AI models perform unevenly across languages and cultural contexts, and students writing in an additional language face both opportunities and hazards: the tool can bridge a linguistic gap, but it can also flatten the author’s voice, introduce subtle factual errors, and produce text that sounds native while lacking depth. The Indonesian case thus functions as a natural laboratory for studying how AI mediation reshapes academic literacy in the Global South, where adoption has been rapid but institutional guidance has lagged. The study’s insistence on empirical, interaction-level evidence provides a model for researchers in other national contexts who want to move beyond attitude surveys and policy commentary.

For educators, the practical takeaways are concrete. Writing instruction should explicitly teach critical evaluation of AI output, including fact-checking generated claims against real sources, recognizing hallucinated citations, and revising machine suggestions for argumentative fit. Assessment should reward visible process, not just polished product. Institutional policy should define acceptable use with clarity and nuance rather than blunt prohibition, which research consistently shows drives use underground without improving judgment. And instructors themselves need professional development, because students’ AI practices will only be as sound as the feedback culture surrounding them. The study concludes that generative AI is not an external threat to academic writing but a new participant in it, and that the task of universities is to cultivate collaborative intelligence: partnerships in which human writers remain the authors of their ideas while machines serve as capable, critically supervised assistants.

Subject of Research: Indonesian university students' interaction with generative AI in academic writing from a collaborative intelligence perspective.

Article Title: Assessing Indonesian university students’ interaction with generative AI in academic writing activities: a collaborative intelligence perspective

Article References: Mulyono, H., Suryoputro, G., Azizah, U. S. A., & Falah, Z. A. (2026). Assessing Indonesian university students’ interaction with generative AI in academic writing activities: a collaborative intelligence perspective. SN Social Sciences, 6(9), Article 425. https://doi.org/10.1007/s43545-026-01716-x

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01716-x

Keywords: generative AI, academic writing, collaborative intelligence, Indonesian university students, human-AI interaction, higher education, AI literacy, assessment, prompting, academic integrity, SN Social Sciences, Assessing