Ethical use of ChatGPT in education—Best practices to combat AI-induced plagiarism

Frontiers in Education, Journal Year: 2025, Volume and Issue: 9

Published: Jan. 9, 2025

The emergence of ChatGPT, a high-performance artificial intelligence language model developed by OpenAI, has generated both excitement and concern in academia (Li, 2024). Equipped with advanced natural processing techniques, ChatGPT is able to generate human-like text that provides coherent contextually relevant responses wide range queries. This unprecedented capability raised optimism as it could fundamentally change traditional practices academia, industry everyday life (Cambra-Fierro et al., 2024).The basic function "ask me anything" "I might have good answer" no longer just many fields. scientific knowledge disseminated journals already struggling the role such technology will play. Questions arise about whether be, can coauthored (Tang, Professors who create immediately face challenge assessing students presence technology. These are practical legitimate questions.While benefits terms increased student engagement, collaboration accessibility outcomes, also very serious academic integrity implications: at its core plagiarism. paper offers comprehensive strategies on how educators help mitigate these risks promoting ethical use fairness within AI tools.ChatGPT was truly disruptive, which should surprised one. It be seen technologies being adopted quickly from university labs; reached one million users first five days now over 180 (Duarte, kind rapid adoption demonstrates remarkable property generative AI: persists text.One main problems models like threats they pose, including black box algorithms, discrimination, biases, vulgarity, copyright infringement -plagiarism -and others, generation fake content or media (Sloan, Powis & Tan, Therefore, organisations need disciplined risk management approaches effectively address threats. Considering continuous evolution algorithms due rapidity data sources, review heterogeneity variability bias periodic assessments weighed against considerations (Schwartz 2022).The experience resulting lacked an obvious logical structure, contained speculative information, did not elaborate critical data, provide original contributions (Giuggioli Pellegrini 2023). Any article topic would conventional, lack logic facts, critically engaging. In addition, references generally incorrect; titles authors, well other publication details, misstated. Such inaccuracies require careful double-checking, especially professional contexts journalism software development.Inaccuracy, poor flow, factual inaccuracies, analysis originality AI-generated result current state (Yang, based deep learning trained using extensive datasets prior information may outdated low quality. Although improvements training quality improve performance AIs, clear technical level necessarily lead significant gains innovation (Dwivedi 2023).The recent applications text, film music production all indicate platforms best partners process, complementing rather than replacing human intelligence. case complex activities requiring creativity emotional intelligence, well-formulated request alone sufficient for produce markedly different outputs. Human oversight remain essential (Liu, Research, practice, urgent policy decisions era rapidly evolving researchers, practitioners, policymakers engage changes. Building strengths AI, while aware limitations making efforts them, foster environment tools used responsibly effectively.Integrating into without challenges. primary possibility Students get essays assignments, then submit their own work. undermines educational process devalues credentials. Another potential inequality. access complete assignments much less time possibly better, giving them unfair advantage do ChatGPT. further increase existing inequalities outcomes. On hand, difficult distinguish created AI. Because generates human-like, difficulty distinguishing "original" makes detect AI-assisted plagiarism.While this work focuses addressing plagiarism, hold great promise improving outcomes stimulating creativity. Through adaptive tutoring systems, personalized learning, immediate feedback facilitate deeper interaction course material. Furthermore, AI-driven creative allow experiment problem-solving thinking new ways, ultimately more dynamic engaging environment.The rise large models, education led institutions develop ways prevent misuse. aim protect adapting AI-enhanced environments. Different been introduced settings varying degrees success.This probably reason why facilities started establish policies when employ Many often tend explain emphasis proper citation attribution student's For example, some universities mention what tool throughout assignment, similar citing sources literature.A number implemented high-tech, AI-detecting plagiarism-checking programs. Indeed, services Turnitin year flagging submissions out character and/or contain unnatural patterns speech. designed AIassisted implemented, complicating misrepresent own.Another effective strategy design increasingly high part student, effective. personal reflective nature, those research questions specific local contexts, make harder fall back only. minimizes only chances misuse but fosters skills among students.Some adopting oral examinations wherein made present defend ideas, projects. face-to-face virtual exchanges permit instructor directly determine depth understanding exams, almost impossible because involves real-time response justification.In where group fostered, teams projects, raises noticeable obstacles fit smoothly inside final product. Group-based nature communication, coordination, team members, aspects imitate. Moreover, mechanisms peer evaluate colleagues, thus automatically increasing identification inconsistencies any tools.Empirical evidence supports importance evaluation reduce AIrelated Successful pilot programs highly regarded colleges incorporate tasks highlighted Moorhouse al (2023). limit individualized tailored students. Dempere (2023) favor technology-based ethics-based interventions, showing campaigns combination recognition greatly compliance. Taken together, studies show integrating awareness assessment strong foundation successful prevention AIenabled plagiarism.To challenges education, Cotton (2024) highlight dual highlighting associated prospect engagement. They call proactive institutional measures integration AIrecognition technologies, creation explicit tools. By implementing tactics, encourage Zeb higher pointing engagement related integrity. recommend implement policies, thinking, guide use. measures, harness minimizing misuse.Strategies taking account opinions suggestions:Technological Solutions• There various plagiarism detectors find copied content. If there search texts match possible flagged. Educators invest intelligence-generated through stylistic anomalies. • Use analytics track learner progress unexplained performance. include sudden, improvement writing style, sign AI-enabled testing methods modified reformulated previous responses. with, predict correct answers incorporating approaches.• Educating most combat education. exactly damage does built name institutions. achieved teaching materials, classroom discussions, communication consequences Include exercises learners discuss encountered, insights gained. teachers assess credibility students' understand processes. Peer incorporated, asked each other's submitted allows detected. Encourage projects pupils individual, products multimedia presentations senses. podcasts digital unlikely replicable AI.• Design linking experiences, curricula. types general addition written essay, communicate learned variety media, slide shows, audio recordings, films, portfolios. replicating alternative methods, versatile skills.• Setting guidelines essential. know context tools, i.e., texts. Requiring outline instructors identify early process. approach timely guidance, reducing likelihood resorting Regularly checking thorough reading check understanding, detection devices flag suspicious content.• Large broken down smaller structured key points, appropriate deadlines. ensures build up gradually, whole project Oral sure test originality; justify arguments even answers, sense scenario.To minimize prone Some extended misuse:Critical Thinking Problem-Solving Tasks• Tasks problem solving performed satisfactorily presentations, interactive individual skills. Designing open-ended conditions useful. having formulate independent thinking. Refine focus areas short, in-depth responses.• Demonstrate applications: apply theoretical practical, real-world problems. Case studies, simulations, project-based AI's ability limited. replicate real-life situations authentic service-learning internships, community-based research. cannot easily outsourced Develop role-playing simulations take designated roles characters. way elements simulate.• Create cohort events, problems, reflections. Individualizing applicability Providing follow-up actions feedback, material reduces reliance portfolio-based assessment, collects done time. Portfolios challenging simulate.Collaborative Peer-Based Learning• Group together product, ensuring input requires coordination replicate. Peer-assisted activities, tutor mentor classmates. reinforces explanation justification, provide.• Real-time proctored exams assessments. represents abilities. Conduct timed assessments, in-class online tests, format emphasizes think respond knowledge.• mixed forms assessment: work, demonstrations.Multimodal diverse skills, handle elements. Interactive vary performance, personalization Frequent, low-level monitor ongoing basis. irregularities last-minute AI.Although above-mentioned tactic sound method curbing application difficulties.Some universities, limited resources, costs sophisticated prohibitive. effectiveness depends frequent updates keep pace capabilities, operational costs.Many solutions, testing, collection amounts data. security privacy, sensitive required. scope collected examined fact schools organizations must ensure compliance protection laws.Authentic mistaken AI-based certain distinctive style. false claims undermine confidence manual investigation teachers, time-and resource-intensive process.Teachers devote lot energy pedagogical policy-based dividing ones. support successfully integrate changes daily routines.A heavy techniques divert attention raising awareness. While resources helpful, remains developing long-lasting moral behavior.Since constantly changing, checked.Institutions regularly adjust evolves, necessitating regulatory teacher training.Administrative faculty burdened change.These discussed coincided globally begun start. next section point similarities between recommendations relative perhaps form prevention. Nothing works better incorrect usage. Institutions really involved culture integrity, taught future careers affected dishonestly passed practices. workshops modules teach ethics mind-reminding attribution.These adaptive, updated performance-real-time quizzes work-are important deterrents growing Adaptive tests responses, quite answers. Continuous approaches-including low-stakes assignments-help students, underlining discrepancies indicative As position follow iterations become vulnerable submissions.Multimodal becoming preferred fighter dishonesty demonstrations raise expectation demonstrate wider assessments-where collect body semester-offer panoramic view development probable spot style.Already, Trialling type Early suggests accuracy continuously systems perception. long-term impacts.Researchers, practitioners makers explore ever-changing technologies. moves direction providing environment.

Language: Английский

---

Using AI chatbots (e.g., CHATGPT) in seeking health-related information online: the case of a common ailment

Computers in Human Behavior Artificial Humans, Journal Year: 2025, Volume and Issue: unknown, P. 100127 - 100127

Published: Feb. 1, 2025

Language: Английский

---

Enhancing Self‐Esteem: Evaluating the Effects of a Self‐Affirmation Intervention Among Indian Adults With Subclinical Depression

Counselling and Psychotherapy Research, Journal Year: 2025, Volume and Issue: 25(1)

Published: Feb. 2, 2025

ABSTRACT Objectives Previous research reports self‐affirmation interventions to be efficacious in enhancing many life outcomes of normal individuals, but limited has assessed its restorative and preserving capacities enhance maintain the self‐esteem individuals with subclinical depressive tendencies. Methods This study employed an experimental design. Eighty participants (age range = 22–27 years) depression were chosen using purposive sampling randomly assigned equal numbers control conditions their was measured at pre‐intervention, post‐intervention follow‐up intervals. Results showed that group evoked significantly higher mean scores ( F (1, 76) 29.20, p < 0.001, η 2 0.28) on as compared (Control: Mean SD ) 17.48 (2.73); Experimental: 51.43 (3.29)) 18.28 (2.72); 48.65 (3.74)). There a significant difference during (2, 152) 6.50, 0.002, 0.08). These findings evident conditions' (experimental, control) intervals' (pre‐intervention, post‐intervention, follow‐up) main interaction effects: 6.54, 0.08; 3.87, 0.023, 0.05. The size these effects ranged from very low, medium ). Conclusion revealed this intervention for self‐esteem. positive may have been produced result activation meaning, cognitions relationships.

Language: Английский

---

Synchronous, Moderated, and Anonymous Peer Support Chats Reduce Momentary Loneliness in Older Adults: A Retrospective Observational Study (Preprint)

JMIR Formative Research, Journal Year: 2024, Volume and Issue: 8, P. e59501 - e59501

Published: Sept. 3, 2024

Background Older adults have a high rate of loneliness, which contributes to increased psychosocial risk, medical morbidity, and mortality. Digital emotional support interventions provide convenient rapid avenue for additional support. peer struggles contrast the usual provider-based clinical care models because they offer more accessible, direct empowerment, highlighting users’ autonomy, competence, relatedness. Objective This study aims examine novel anonymous synchronous peer-to-peer digital chat service facilitated by trained human moderators. The experience cohort 699 aged ≥65 years was analyzed determine (1) if participation, alone, led measurable aggregate change in momentary loneliness optimism (2) impact peers on optimism. Methods Participants were each prompted with single question: “What’s your struggle?” Using proprietary artificial intelligence model, free-text response automatched respondent based their self-expressed struggle moderator. Exchanged messages quantitatively measure using third-party, public, natural language processing model (GPT-4 [OpenAI]). sentiment analysis initially performed at individual level then averaged across all users similar emotion types produce statistically significant (P<.05) collective trend per emotion. To evaluate optimism, we propensity matching align moderator+single user moderator+small group cohorts compare trends between matched cohorts. Results Loneliness significantly improved after 8 (P=.02) 9 minutes (P=.03) into chat. We observed improvement compared 19 (P=.049) 21 (P=.04) respectively. Conclusions Chat-based may be viable intervention help address older present an alternative traditional care. promising results need further expand evidence such cost-effective options.

Language: Английский

---

Pilot Study on AI Image Analysis for Lower-Limb Reconstruction—Assessing ChatGPT-4’s Recommendations in Comparison to Board-Certified Plastic Surgeons and Resident Physicians

Life, Journal Year: 2025, Volume and Issue: 15(1), P. 66 - 66

Published: Jan. 8, 2025

AI, especially ChatGPT, is impacting healthcare through applications in research, patient communication, and training. To our knowledge, this the first study to examine ChatGPT-4's ability analyze images of lower leg defects assesses its understanding complex case reports comparison performance board-certified surgeons residents. We conducted a cross-sectional survey Switzerland, Germany, Austria, where 52 participants reviewed depicting within fictitious profiles selected optimal reconstruction techniques. The questionnaire included cases with varied difficulty, answer options did not always include most obvious choices. Findings highlight that ChatGPT-4 successfully evaluated various methods but struggled determine solution based on available information visual written forms. A chi-squared test independence was performed investigate overall association between (A, B, C, D) rater group (board-certified surgeons, ChatGPT-4, resident). Inter-group associations showed significant results (p < 0.001), high agreement among surgeons. Our suggest plastic remain essential for patient-specific treatment planning, while AI can support decision-making. This reaffirms role as supportive tool, rather than replacement, reconstructive surgery.

Language: Английский

---

Advancing Emotionally Aware Child–Robot Interaction with Biophysical Data and Insight-Driven Affective Computing

Sensors, Journal Year: 2025, Volume and Issue: 25(4), P. 1161 - 1161

Published: Feb. 14, 2025

This paper investigates the integration of affective computing techniques using biophysical data to advance emotionally aware machines and enhance child-robot interaction (CRI). By leveraging interdisciplinary insights from neuroscience, psychology, artificial intelligence, study focuses on creating adaptive, emotion-aware systems capable dynamically recognizing responding human emotional states. Through a real-world CRI pilot involving NAO robot, this research demonstrates how facial expression analysis speech emotion recognition can be employed detect address negative emotions in real time, fostering positive engagement. The system combines handcrafted deep learning features for expressions, achieving an 85% classification accuracy during real-time CRI, while are analyzed acoustic processed through machine models with 83% rate. Offline evaluation combined dataset Dynamic Bayesian Mixture Model (DBMM) achieved 92% multilingual yielded 98% DBMM ensemble. Observations psychological technological aspects, coupled statistical analysis, reveal robot's ability transition into neutral or states most cases, contributing regulation children. work underscores potential robots support therapeutic educational interventions, particularly pediatric populations, setting foundation developing personalized empathetic human-machine interactions. These findings demonstrate transformative role bridging gap between functionality intelligence across diverse domains.

Language: Английский

---

Hysteria in empathy: Understanding virtual companionship and emotional connection between humans and Al

Theory & Psychology, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 24, 2025

Increasing numbers of people tend to seek relationships with artificial intelligence (AI) because its comprehensive mastery knowledge. This research reveals that ChatGPT has become a temporary refuge for individuals encountering real-life emotional crises. The communicative affordance human–AI conversation is featured by logos-centrism, pursuing the identity language and essence between humans technology. Grounded in Lacanian discourse theory, this argues psychological discursive dynamics behind dialogue entail structure hysteric’s discourse, which stems from questioning users fantasizing interrogating GPT as master Also, capitalist revealed during repetitive AI companionship empowers agency controlling user’s primordial lack desire transformed into demand can be resolved through consumption. may consumer product technological capital, while needs are commodified within these structures.

Language: Английский

---

Integrating ChatGPT in halal tourism: impact on tourist satisfaction, e-WoM and revisit intention

Journal of Islamic marketing, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 28, 2025

Purpose This study aims to examine the impact of using ChatGPT on Halal tourism experience. It examines relationships among Halal-friendly travel motivations and satisfaction, revisit intention electronic word-of-mouth (e-WoM) while testing moderating effect relationship between satisfaction intention. Design/methodology/approach employed a quantitative methodology. Using purposive sampling techniques, it approached about 800 tourists (from November 2023 January 2024) from several halal destinations in Indonesia. A total 395 usable surveys were analyzed test moderation effects by SEM. Findings The indicates that positively Muslim tourist which turn influences e-WoM Importantly, significantly moderates intention, thereby strengthening loyalty for those AI tool. Practical implications study’s findings provide practical guidelines providers enhance Halal-compliant services incorporate as an tool boost travelers’ drive increase intentions. technology gives companies advantage offering customized, immediate support, leads visitors becoming loyal. Originality/value fills significant gap literature examining AI’s market. expands Expectation-Confirmation Theory (ECT), push-pull theory models tourism. also contributes adoption addressing how modern tools can influence behaviors, improve encourage repeat visits.

Language: Английский

---

Adaptive human-computer interaction for industry 5.0: A novel concept, with comprehensive review and empirical validation

Computers in Industry, Journal Year: 2025, Volume and Issue: 168, P. 104268 - 104268

Published: March 1, 2025

Language: Английский

---

The Role of Large Language Model Chatbots in Sexual Education: An Unmet Need of Research

Journal of Psychosexual Health, Journal Year: 2025, Volume and Issue: unknown

Published: March 17, 2025

Large language model (LLM) chatbots have demonstrated significant capability in patient education by offering accessible, consistent, and personalized information. Their ability to interact real-time adapt responses based on user input makes them valuable tools enhancing knowledge engagement. Sexual developing countries faces substantial challenges. Sociocultural barriers, limited access comprehensive educational resources, stigmatization surrounding sexual health contribute inadequate education. Traditional methods often fail reach remote or underserved populations, there is a general shortage of qualified educators resources. Chatbots present promising solution these They can offer anonymous, culturally sensitive information health, overcoming barriers related stigma privacy. While LLM hold potential improve countries, their implementation must be carefully managed address challenges such as ensuring accuracy cultural sensitivity. There dearth research Hence, unmet need the reliability information, maintaining sensitivity, assessing engagement, integration with traditional methods, exploring long-term impact improving knowledge.

Language: Английский

---