L’éthique de l’intelligence artificielle (IA) constitue un domaine de recherche pluridisciplinaire en expansion. Cette thèse s’inscrit dans le champ de l’éthique de l’IA et en propose une nouvelle interprétation. D’entrée de jeu, les nombreux liens qu’entretiennent les défis et opportunités que présentent les systèmes employant l’intelligence artificielle avec le domaine politique y sont exposés. Une hypothèse clé animant cette thèse est que, puisque la politique consiste à répondre au conflit par le dialogue, les décideurs politiques aux prises avec des questions concernant l’IA peuvent tirer profit des orientations fournies par différentes traditions éthiques. Afin de faciliter un dialogue optimal, une prise de position métaéthique particulière est avancée au niveau théorique, soit le « monisme non orthodoxe », tandis qu’une série de questions ciblées est proposée au niveau de la pratique. Divisée en trois sections, la thèse débute avec une exploration métaéthique des fondements des approches éthiques principales qui sont à l’œuvre dans les réflexions contemporaines concernant l’IA. Les écoles éthiques étudiées, en faisant appel au continuum qui distingue le monisme et le pluralisme, sont l’éthique de la vertu, l’utilitarisme et l’éthique déontologique. Ces démarches monistes sont ensuite placées en contraste avec le pluralisme des valeurs, une approche souvent employée, mais rarement nommée de manière explicite. La deuxième section consiste en une analyse métaéthique d’une vingtaine de directives éthiques émises par des compagnies privées, la société civile ainsi que des organisations à multiples partenaires, de même que par des instances gouvernementales ou intergouvernementales. C’est ce portrait qui révèle à quel point le pluralisme des valeurs est récurrent dans ces directives. En outre, il se mêle souvent à d’autres approches éthiques pour générer des versions en situation de « tension métaéthique », bien que cela ne se produise souvent que de manière implicite. En conséquence, les propositions associées à ces approches sont parfois contradictoires, tant en ce qui concerne leur formulation que dans la manière dont elles seraient mises en œuvre. Une approche éthique alternative est proposée dans la troisième section. Cette approche est formée d’éléments dérivés spécialement de l’éthique de la vertu et du pluralisme. Ils fondent le socle sur lequel, dans le chapitre final de la thèse, un « guide dialogique » est développé pour l’usage des décideurs politiques. La prudence, une sensibilité profonde au contexte, une orientation téléologique « douce » vers le bien commun ainsi que l’ouverture à la possibilité de dilemmes insolubles caractérisent cette approche éthique. La philosophie herméneutique est également mise à contribution pour justifier l’articulation d’une série de questions destinées à guider le dialogue des décideurs politiques. En effet, l’herméneutique encourage une logique de questions, plutôt qu’une logique dérivée de principes ou de théories. / The ethics of artificial intelligence (AI) is a growing field of multidisciplinary research. This thesis falls within AI ethics and suggests a new interpretation. The numerous challenges and opportunities generated by AI systems are described at the outset. A key assumption of the thesis is that, given that politics consists of responding to conflict with dialogue, policy-makers dealing with the questions surrounding AI can benefit from the guidance provided by different ethical traditions. Furthermore, in order to facilitate an optimal dialogue, this thesis puts forward a particular metaethical position, that of “unorthodox monism” at the theoretical level, and one consisting of a series of pertinent questions at the practical level. Divided into three main sections, the thesis begins by exploring the metaethical foundations at work in most contemporary thinking about AI. Using a continuum that distinguishes between monism and pluralism in ethics, the particular schools examined are virtue ethics, utilitarianism, and deontology. These monistic approaches are then contrasted with value pluralism, which is an approach that is often employed and yet rarely identified explicitly. The second section consists of a metaethical analysis of a sample of some twenty sets of AI ethical guidelines produced within private companies, civil society and by multi-partner organizations, as well as by governmental and intergovernmental bodies. It is this portrait which reveals how value pluralism recurs in many of these sets of guidelines. Moreover, it is also often implicitly combined with other ethical approaches to generate “mixed” versions, which exhibit metaethical tensions. As a result, the proposals associated with these approaches are sometimes contradictory, both in regards to how they are formulated as well as to how they are to be implemented. Third, an alternative ethical approach is proposed. It consists of elements derived in particular from virtue ethics and value pluralism. They form the basis upon which, in the thesis’ final chapter, a “dialogical guide” is formulated for use by policy-makers. Prudence, a deep sensitivity to context, a “soft” teleological orientation towards a common good, along with openness to the possibility of intractable dilemmas characterize this approach. Hermeneutical philosophy is also drawn upon in order to justify the articulation of a series of questions meant to guide the dialogue of policy-makers. Indeed, hermeneutics calls forth a logic of questions rather than one based on principles or theories.

Supply chain operations of construction industry including road projects in disaster regions results in exceeding project budget and timelines. In road construction projects, supply chain with poor performance can affect efficiency and completion time of the project. This is also the case of the road projects in disaster areas. Disaster areas consider both natural and man-made disasters. Few examples of disaster zones are; Pakistan, Afghanistan, Iraq, Sri Lanka, India, Japan, Haiti and many other countries with similar environments. The key factors affecting project performance and execution are insecurity, uncertainties in demand and supply, poor communication and technology, poor infrastructure, lack of political and government will, unmotivated organizational staff, restricted accessibility to construction materials, legal hitches, multiple challenges of hiring labour force and exponential construction rates due to high risk environment along with multiple other factors. The managers at all tiers are facing challenges of overrunning time and budget of supply chain operations during planning as well as execution phase of development projects. The aim of research is to develop a Multiple Linear Regression Model (MLRM) and a Rule Based Decision Support System by incorporating various factors affecting supply chain management of road projects in disaster areas in the order of importance. This knowledge base (KB) (importance / coefficient of each factor) will assist infrastructure managers (road projects) and practitioners in disaster regions in decision making to minimize the effect of each factor which will further help them in project improvement. Conduct of Literature Review in the fields of disaster areas, supply chain operational environments of road project, statistical techniques, Artificial Intelligence (AI) and types of research approaches has provided deep insights to the researchers. An initial questionnaire was developed and distributed amongst participants as pilot project and consequently results were analysed. The results’ analysis enabled the researcher to extract key variables impacting supply chain performance of road project. The results of questionnaire analysis will facilitate development of Multiple Linear Regression Model, which will eventually be verified and validated with real data from actual environments. The development of Multiple Linear Regression Model and Rule Based Decision Support System incorporating all factors which affect supply chain performance of road projects in disastrous regions is the most vital contribution to the research. The significance and novelty of this research is the methodology developed that is the integration of those different methods which will be employed to measure the SCM performance of road projects in disaster areas.

Eftersom samhället blir allt mer datadrivet revolutionerar användningen av AI och maskininlärning sättet företag fungerar och utvecklas på. Denna studie utforskar användningen av AI, Big Data och Natural Language Processing (NLP) för att förbättra affärsverksamhet och intelligens i företag. Huvudsyftet med denna avhandling är att undersöka om den nuvarande klassificeringsprocessen hos värdorganisationen kan upprätthållas med minskade driftskostnader, särskilt lägre moln-GPU-kostnader. Detta har potential att förbättra klassificeringsmetoden, förbättra produkten som företaget erbjuder sina kunder på grund av ökad klassificeringsnoggrannhet och stärka deras värdeerbjudande. Vidare utvärderas tre tillvägagångssätt mot varandra och implementationerna visar utvecklingen inom området. Modellerna som jämförs i denna studie inkluderar traditionella maskininlärningsmetoder som Support Vector Machine (SVM) och Logistisk Regression, tillsammans med state-of-the-art transformermodeller som BERT, både Pre-Trained och Fine-Tuned. Artikeln visar att det finns en avvägning mellan prestanda och kostnad vilket illustrerar problemet som många företag, som Valu8, står inför när de utvärderar vilket tillvägagångssätt de ska implementera. Denna avvägning diskuteras och analyseras sedan mer detaljerat för att utforska möjliga kompromisser från varje perspektiv i ett försök att hitta en balanserad lösning som kombinerar prestandaeffektivitet och kostnadseffektivitet. / As society is becoming more data-driven, Artificial Intelligence (AI) and Machine Learning are revolutionizing how companies operate and evolve. This study explores the use of AI, Big Data, and Natural Language Processing (NLP) in improving business operations and intelligence in enterprises. The primary objective of this thesis is to examine if the current classification process at the host company can be maintained with reduced operating costs, specifically lower cloud GPU costs. This can improve the classification method, enhance the product the company offers its customers due to increased classification accuracy, and strengthen its value proposition. Furthermore, three approaches are evaluated against each other, and the implementations showcase the evolution within the field. The models compared in this study include traditional machine learning methods such as Support Vector Machine (SVM) and Logistic Regression, alongside state-of-the-art transformer models like BERT, both Pre-Trained and Fine-Tuned. The paper shows a trade-off between performance and cost, showcasing the problem many companies like Valu8 stand before when evaluating which approach to implement. This trade-off is discussed and analyzed in further detail to explore possible compromises from each perspective to strike a balanced solution that combines performance efficiency and cost-effectiveness.

Air quality assessment plays a crucial role in understanding the impact of air pollution onhuman health and the environment. With the increasing demand for accurate assessment andprediction of air quality, advanced data analytics modelling techniques offer promisingsolutions. This thesis focuses on leveraging advanced data analytics to assess and analyse airpollution concentration levels in Italy over a 4km resolution using the FORAIR_IT datasetsimulated in ENEA on the CRESCO6 infrastructure, aiming to uncover valuable insights andidentifying the most appropriate AI models for predicting air pollution levels. The datacollection, understanding, and pre-processing procedures are discussed, followed by theapplication of big data training and forecasting using Apache Spark MLlib. The research alsoencompasses different phases, including descriptive and inferential analysis to understand theair pollution concentration dataset, hypothesis testing to examine the relationship betweenvarious pollutants, machine learning prediction using several regression models and anensemble machine learning approach and time series analysis on the entire dataset as well asthree major regions in Italy (Northern Italy – Lombardy, Central Italy – Lazio and SouthernItaly – Campania). The computation time for these regression models are also evaluated and acomparative analysis is done on the results obtained. The evaluation process and theexperimental setup involve the usage of the ENEAGRID/CRESCO6 HPC Infrastructure andApache Spark. This research has provided valuable insights into understanding air pollutionpatterns and improving prediction accuracy. The findings of this study have the potential todrive positive change in environmental management and decision-making processes, ultimatelyleading to healthier and more sustainable communities. As we continue to explore the vastpossibilities offered by advanced data analytics, this research serves as a foundation for futureadvancements in air quality assessment in Italy and the models are transferable to other regionsand provinces in Italy, paving the way for a cleaner and greener future.

Background: Capabilities of Artificial Intelligence (AI) are rapidly advancing, as are its potential applications. Examples of the adoption of AI in social work already exist, but an overview of its manifold uses is lacking. This review aimed to systematically assess the existing research focused on the uses of AI applications in social work practice and to spotlight use-cases yet to be explored. Methods: A scoping review was conducted guided by Arksey and O'Malley's framework and adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis, extension for Scoping Review (PRISMA-ScR). A systematic search was performed using the Scopus database. Eligibility criteria included pre-prints and published articles from January 2000 to April 2023 that emphasized AI implementations in social work practice. No limitations were placed on study design. Data extracted included: article details; country of study; the AI use-case and task; and the specific AI technology employed. Extracted data from all eligible studies were collated using tables and accompanied by narrative descriptive summaries. The review employed CAIMeR (a theory explaining the results of social work interventions) to  pinpoint gaps and highlight novel unexplored applications of AI in social work.  Results: Of the 159 identified articles, 28 satisfied the inclusion criteria. On average, three relevant publications surfaced annually, with approximately 60% hailing from the US. Notably, the absolute majority of the applications of AI were concentrated on predicting or elucidating individual’s health or social condition. Conclusion: Although AI possesses substantial potential, current research into its applications in social work remains surprisingly sparse and averaging a mere three studies annually. The prevailing emphasis of this research is on discerning individual health or social conditions. Given AI's multifaceted capabilities, there exists a substantial opportunity to broaden research into other applications. Informed by the CAIMeR theory, this review identifies several unexplored applications of AI paving the way for future research. / Bakgrund: Utvecklingen inom Artificiell Intelligens (AI) medför betydande potentiella fördelar och utmaningar, vilket understryker behovet för det socialt arbetets praktik att anpassa och ta till sig dess användning. Denna studie undersöker användningen av AI inom socialt arbete genom att kartlägga inom vilka domäner av socialt arbete AI har använts och för vilket syfte. Därtill identifieras forskningsluckor och nya användningsområden för AI med hjälp av CAIMeR teorin. Metod: Genom att använda en scoping review metodik vägledd av Arksey och O'Malleys ramverk och PRISMA-ScR:s riktlinjer, utfördes en systematisk sökning i Scopus fram till april 2023 med fokus på artiklar som diskuterar AI:s implementering i socialt arbete. Resultat: Av 159 artiklar som hittades uppfyllde 28 inkluderingskriterierna. AI har använts flitigt inom socialt arbete, främst för att förutsäga eller diagnostisera individers tillstånd. Forskningsvolymen är begränsad, med ungefär tre studier som genomförts årligen. Slutsats: Trots AI:s potential att förbättra socialt arbete visar nuvarande litteratur en begränsad forskningsvolym om ämnet och ett begränsat användningssätt för AI. Nästan uteslutande koncentrerar sig studierna på användningen av AI för att förutsäga sociala problem eller hälsotillstånd. Studien identifierar ett behov av att utforska AI inom flera användningsområden inom socialt arbete. Med hjälp av CAIMeR-teorin presenterar denna studie flera sådana potentiella användningsområden av AI.

<p>  </p> <p>Traditional manual code compliance checking process is a time-consuming, costly, and error-prone process that has many shortcomings (Zhang & El-Gohary, 2015). Therefore, automated code compliance checking systems have emerged as an alternative to traditional code compliance checking. However, computer software cannot directly process regulatory information in unstructured building code texts. To support automated code compliance checking, building codes need to be transformed to a computer-processable, structured format. In particular, the problem that most automated code compliance checking systems can only check a limited number of building code requirements stands out.</p> <p>The transformation of building code requirements into a computer-processable, structured format is a natural language processing (NLP) task that requires highly accurate part-of-speech (POS) tagging results on building codes beyond the state of the art. To address this need, this dissertation research was conducted to provide a method to improve the performance of POS taggers by error-driven transformational rules that revise machine-tagged POS results. The proposed error-driven transformational rules fix errors in POS tagging results in two steps. First, error-driven transformational rules locate errors in POS tagging by their context. Second, error-driven transformational rules replace the erroneous POS tag with the correct POS tag that is stored in the rule. A dataset of POS tagged building codes, namely the Part-of-Speech Tagged Building Codes (PTBC) dataset (Xue & Zhang, 2019), was published in the Purdue University Research Repository (PURR). Testing on the dataset illustrated that the method corrected 71.00% of errors in POS tagging results for building codes. As a result, the POS tagging accuracy on building codes was increased from 89.13% to 96.85%.</p> <p>This dissertation research was conducted to provide a new POS tagger that is tailored to building codes. The proposed POS tagger utilized neural network models and error-driven transformational rules. The neural network model contained a pre-trained model and one or more trainable neural layers. The neural network model was trained and fine-tuned on the PTBC (Xue & Zhang, 2019) dataset, which was published in the Purdue University Research Repository (PURR). In this dissertation research, a high-performance POS tagger for building codes using one bidirectional Long-short Term Memory (LSTM) Recurrent Neural Network (RNN) trainable layer, a BERT-Cased-Base pre-trained model, and 50 epochs of training was discovered. This model achieved 91.89% precision without error-driven transformational rules and 95.11% precision with error-driven transformational rules, outperforming the otherwise most advanced POS tagger’s 89.82% precision on building codes in the state of the art.</p> <p>Other automated information extraction methods were also developed in this dissertation. Some automated code compliance checking systems represented building codes in logic clauses and used pattern matching-based rules to convert building codes from natural language text to logic clauses (Zhang & El-Gohary 2017). A ruleset expansion method that can expand the range of checkable building codes of such automated code compliance checking systems by expanding their pattern matching-based ruleset was developed in this dissertation research. The ruleset expansion method can guarantee: (1) the ruleset’s backward compatibility with the building codes that the ruleset was already able to process, and (2) forward compatibility with building codes that the ruleset may need to process in the future. The ruleset expansion method was validated on Chapters 5 and 10 of the International Building Code 2015 (IBC 2015). The Chapter 10 of IBC 2015 was used as the training dataset and the Chapter 5 of the IBC 2015 was used as the testing dataset. A gold standard of logic clauses was published in the Logic Clause Representation of Building Codes (LCRBC) dataset (Xue & Zhang, 2021). Expanded pattern matching-based rules were published in the dissertation (Appendix A). The expanded ruleset increased the precision, recall, and f1-score of the logic clause generation at the predicate-level by 10.44%, 25.72%, and 18.02%, to 95.17%, 96.60%, and 95.88%, comparing to the baseline ruleset, respectively. </p> <p>Most of the existing automated code compliance checking research focused on checking regulatory information that was stored in textual format in building code in text. However, a comprehensive automated code compliance checking process should be able to check regulatory information stored in other parts, such as, tables. Therefore, this dissertation research was conducted to provide a semi-automated information extraction and transformation method for tabular information processing in building codes. The proposed method can semi-automatically detect the layouts of tables and store the extracted information of a table in a database. Automated code compliance checking systems can then query the database for regulatory information in the corresponding table. The algorithm’s initial implementation accurately processed 91.67 % of the tables in the testing dataset composed of tables in Chapter 10 of IBC  2015. After iterative upgrades, the updated method correctly processed all tables in the testing dataset. </p>

The pharmaceutical sector is important for human health due to the increasing demand for medical products but is also a pollution and waste intensive industry. There is an urgent need for the industry to review its environmental footprints and simultaneously consider the industrial transformation called Industry 4.0. This is especially true for small and medium sized enterprises (SMEs). To achieve these objectives, it is presumed that artificial intelligence (AI) will have an important role.  This thesis sets out to identify barriers that pharmaceutical SMEs may encounter when implementing AI to improve environmental efforts. Furthermore, due to the lack of efficient tools the Green Performance Map is analyzed to see if additional value in the design phase and running of a production system can be obtained. Semi-structured interviews were conducted as this thesis is a case study and follows an inductive process. Other qualitative data and literature were used to investigate the research questions. The results indicate that organizational, resource, regulatory and knowledge specific factors can create barriers. Furthermore, there are indications that the Green Performance Map will be useful in both the design phase and running of a production system, this is however dependent on the resources. / Läkemedelssektorn är viktig för människans hälsa på grund av den ökade efterfrågan av medicinska produkter men bidrar även till stora mängder avfall och föroreningar. Det finns ett akut behov att industrin granskar sina miljöavtryck och samtidigt överväger den industriella omvandlingen som kallas Industri 4.0. Detta gäller särskilt för små och medelstora företag. För att uppnå dessa mål kan Artificiell Intelligens (AI) komma att ha en betydelsefull roll.  Detta examensarbete syftar till att identifiera de hinder som små och medelstora läkemedelsföretag kan stöta på när de implementerar AI för att förbättra sitt miljöarbete. På grund av brist på effektiva verktyg analyseras dessutom Green Performance Map för att se om ett mervärde i designfasen och under driften av produktionssystemet kan erhållas. Semistrukturerade intervjuer genomfördes då examensarbetet är en fallstudie och följer en induktiv process. Kvalitativa data och litteratur användes för att undersöka forskningsfrågorna. Resultatet indikerar att organisatoriska, resurs-, reglerings- och kunskapsspecifika faktorer kan skapa barriärer. Dessutom, finns det indikatorer på att Green Performance Map kommer vara användbart både i designfasen och när produktionssystemet är i drift, detta är dock beroende på nivån av resurser.

The pharmaceutical sector is important for human health due to the increasing demand for medical products but is also a pollution and waste intensive industry. There is an urgent need for the industry to review its environmental footprints and simultaneously consider the industrial transformation called Industry 4.0. This is especially true for small and medium sized enterprises (SMEs). To achieve these objectives, it is presumed that artificial intelligence (AI) will have an important role.  This thesis sets out to identify barriers that pharmaceutical SMEs may encounter when implementing AI to improve environmental efforts. Furthermore, due to the lack of efficient tools the Green Performance Map is analyzed to see if additional value in the design phase and running of a production system can be obtained. Semi-structured interviews were conducted as this thesis is a case study and follows an inductive process. Other qualitative data and literature were used to investigate the research questions. The results indicate that organizational, resource, regulatory and knowledge specific factors can create barriers. Furthermore, there are indications that the Green Performance Map will be useful in both the design phase and running of a production system, this is however dependent on the resources. / Läkemedelssektorn är viktig för människans hälsa på grund av den ökade efterfrågan av medicinska produkter men bidrar även till stora mängder avfall och föroreningar. Det finns ett akut behov att industrin granskar sina miljöavtryck och samtidigt överväger den industriella omvandlingen som kallas Industri 4.0. Detta gäller särskilt för små och medelstora företag. För att uppnå dessa mål kan Artificiell Intelligens (AI) komma att ha en betydelsefull roll.  Detta examensarbete syftar till att identifiera de hinder som små och medelstora läkemedelsföretag kan stöta på när de implementerar AI för att förbättra sitt miljöarbete. På grund av brist på effektiva verktyg analyseras dessutom Green Performance Map för att se om ett mervärde i designfasen och under driften av produktionssystemet kan erhållas. Semistrukturerade intervjuer genomfördes då examensarbetet är en fallstudie och följer en induktiv process. Kvalitativa data och litteratur användes för att undersöka forskningsfrågorna. Resultatet indikerar att organisatoriska, resurs-, reglerings- och kunskapsspecifika faktorer kan skapa barriärer. Dessutom, finns det indikatorer på att Green Performance Map kommer vara användbart både i designfasen och när produktionssystemet är i drift, detta är dock beroende på nivån av resurser.

This research evaluates the technological, organizational, and environmental determinants of emerging technologies adoption represented by Artificial Intelligence (AI) and Internet of Things (IoT) driving SMEs’ competitive advantage within a resource-based view (RBV) theoretical approach supported by the technological-organizational-environmental (TOE)-framework setting. Current literature on SMEs competitive advantage as outcome of emerging technologies in the technological, organisational, and environmental contexts presents models focused on these contexts individual components. There are no models in the literature to represent the TOE framework as an integrated structure with gradual levels of complexity, allowing for incremental evaluation of the business context in support of decision making towards emerging technologies adoption supporting the firm competitive advantage. This research gap is addressed with the introduction of a new concept, the IT resource-based renewal, underpinned by the RBV, and supported by the TOE framework for providing a holistic understanding of the SMEs strategic renewal decision through information technology. This is achieved through a complex measurement model with four level constructs, leading into a parsimonious structural model that evaluates the relationships between IT resource-based renewal, and emerging technologies adoption driving SMEs competitive advantage. The model confirms the positive association between the IT resource-based renewal and emerging technologies adoption, and between the IT resource-based renewal and SME competitive advantage for the SMEs managers model, with the SME owners model outcomes are found not being supportive towards emerging technologies adoption driving SME competitive advantage. As methodology, PLS-SEM is used for its capabilities of assessing complex paths among model variables. Analysis is done on three models, one for the full sample, with two subsequent ones for owners and managers, respectively, as SME decision makers, with data collected using a web-based survey in Canada, the UK, and the US, that has provided 510 usable answers. This research has a theoretical contribution represented by the introduction of the IT resource-based renewal concept, that integrates the RBV perspective and the TOE framework for supporting organization’s decision on emerging technologies adoption driving SMEs competitive advantage. As practical implications, this thesis provides SMEs with a reference framework on adopting emerging technologies, offering SME managers and owners a comprehensive model of hierarchical factors contributing to SMEs competitive advantage acquired as outcome of AI and IoT adoption. This research makes an original contribution to the enterprise management, information systems adoption, and SME competitive advantage literature, with an empirical approach that verifies a model of emerging technologies adoption determinants driving SMEs competitive advantage.

This study investigates the growing integration of Artificial Intelligence (AI) in educational settings, specifically focusing on detecting AI-generated content in students’ English essays. As AI technologies like ChatGPT and Gemini become more prevalent, understanding their impact on education is crucial. This research aims to identify the linguistic features that lead English as a Foreign Language (EFL) teachers to suspect AI involvement in student work. By conducting semi-structured interviews with eight EFL teachers from lower upper secondary and high schools, the study examines their experiences and perspectives. Using the Technological Pedagogical Content Knowledge (TPCK) framework, the study analyzes the crossing of technology, pedagogy, and content knowledge, highlighting the opportunities and challenges AI presents in contemporary education. The findings show the dual role of AI as both a beneficial tool for improving learning and a challenge to maintaining academic integrity. Despite the limitations, such as the evolving nature of AI, the research highlights the need for teachers to balance the benefits of AI with preserving authentic student work. Future research directions include exploring more effective AI detection methods and understanding the long-term impact of AI on students’ critical thinking skills.