Human Oversight Research Articles

TDS-Net: Transformer enhanced dual-stream network for video Anomaly Detection

Video anomaly detection is a critical research area, driven by the increasing reliance on surveillance systems to maintain public safety and security. The implementation of surveillance cameras in public spaces is driven by the specific goal of monitoring incidents involving theft, vandalism, traffic accidents, etc. However, the reliance on human oversight for anomaly detection introduces a susceptibility to errors, emphasizing the urgent need for the development of efficient algorithms capable of autonomously detecting anomalies in video footage. Furthermore, anomalous patterns within video data often have a complex background, unexpected events, variations in the scale of the objects, and anomalous appearances, which makes their detection challenging. Therefore we introduce a novel Dual Stream Transformer Network (TDS-Net) to address these challenges. TDS-Net concurrently extracts RGB and flow features, using a transformer network for sequential pattern learning. The parallel extraction of RGB features allows the model to learn patterns, while the dedicated flow module adeptly handles motion features. The significance of this innovative network for video anomaly detection is that it enhances the ability to identify and analyze unusual patterns or events within video data. Furthermore, the TDS-Net has demonstrated superior accuracy compared to baseline methods across benchmark datasets, including ShanghaiTech, CUHK Avenue, and UCF Crime. We also carried out an ablation study, offering insights into the individual contributions of the components of the TDS-Net architecture.

Legal Horizons of the New Artificial Intelligence Paradigm

Modern society is undergoing a structural transformation of the world economy. This is as a result of the transition to a new technological base through the introduction of artificial intelligence, cutting-edge information and communication technology, energy technology, biotechnology and nanotechnology. Artificial intelligence has the ability to significantly change the economy and social relations in society, and its newly discovered capabilities are transformational and global in nature. At the same time, the extraordinary capabilities of artificial intelligence technologies involve risks that can threaten stability and undermine human values. In order to eliminate possible threats and risks and mitigate potential dangers, it is crucial to develop systemic legal measures and ways to regulate AI technologies and models on a national and international scale and to define the legal status of AI, which must include protection of humans from the uncontrolled influence of AI and the inviolability of guarantees of human rights and freedoms. With this in mind, and in order to mitigate potential dangers and ensure the controllability and sustainability of AI technologies based on the concept of trusted (responsible) AI, it is necessary to agree on universal international guidelines for the development and application of AI technologies and models. Furthermore, it is necessary to create a universal code of conduct for AI developers, who together can create a basis for a uniform framework of legal regulation within the national legislation of each country on the principles of human rights protection, privacy and data protection, transparency and explainability, fairness, accountability and safety of artificial intelligence, adequate human oversight and ethical standards for the creation and application of AI models.

KNOWLEDGE, ATTITUDE, AND PRACTICE REGARDING ARTIFICIAL INTELLIGENCE (AI) AND ITS USAGE IN DENTAL ACADEMICS CURRICULUM AMONG DENTAL UNDERGRADUATES AND POSTGRADUATES

Background: Artificial intelligence (AI) is increasingly being recognized for its potential to revolutionize dental education and practice. However, the extent of knowledge, attitudes, and practices (KAP) regarding AI among dental undergraduates and postgraduates remains underexplored. Objectives: This study aims to assess the knowledge, attitudes, and practices related to AI and its integration into the dental academic curriculum among dental undergraduates and postgraduates. Methods: A cross-sectional questionnaire-based study was conducted among 545 participants, including final-year dental undergraduates and postgraduate dental students . The structured questionnaire covered demographics, knowledge of AI, attitudes towards its integration into dental education, and current practices involving AI. Results: The study found significant variations in knowledge levels, with postgraduates demonstrating a more comprehensive understanding of AI applications in dentistry compared to undergraduates. Attitudes towards AI were generally positive across both groups, with participants recognizing its potential to enhance diagnostic accuracy and streamline administrative tasks. However, there were concerns about the reliability of AI systems and the potential for reduced human oversight. Practical application of AI was limited, particularly among undergraduates, highlighting a gap between theoretical knowledge and practical use. Conclusion: The study underscores the need for integrating comprehensive AI education into the dental curriculum, encompassing both theoretical knowledge and practical training. Continuous professional development and a balanced approach to AI use in dental practice are essential to prepare future dental professionals for an AI-driven healthcare landscape. Gender-wise, the survey reveals consistent support across male and female dental professionals for AI's potential in enhancing dental research, with minimal variation in beliefs. KEYWORDS: Artificial intelligence, dental education, knowledge, attitudes, practices, dental undergraduates, postgraduates.

Ethical Implications of AI in Criminal Justice: Balancing Efficiency and Due Process

The integration of Artificial Intelligence (AI) into criminal justice systems presents a complex landscape of opportunities and challenges. This research article critically examines the ethical implications of AI applications in various facets of criminal justice, from predictive policing to sentencing algorithms. While AI promises enhanced efficiency and data-driven decision-making, it simultaneously raises significant concerns about fairness, transparency, and the preservation of due process rights. This study employs a multidisciplinary approach, combining legal analysis, ethical philosophy, and empirical research to evaluate the current state of AI in criminal justice and project future trajectories. We analyze case studies from multiple jurisdictions, highlighting both successful implementations and controversial failures of AI systems in law enforcement and judicial processes. Our findings reveal a tension between the pursuit of efficiency through AI and the fundamental principles of due process. We identify several key ethical challenges, including algorithmic bias, lack of transparency in decision-making processes, and the potential erosion of human judgment in critical legal determinations. The research also uncovers disparate impacts on marginalized communities, raising concerns about the exacerbation of existing inequalities within the justice system. To address these challenges, we propose a comprehensive ethical framework for the development and deployment of AI in criminal justice. This framework emphasizes the need for continuous human oversight, regular audits of AI systems, and the establishment of clear accountability mechanisms. We argue for a balanced approach that leverages the benefits of AI while safeguarding individual rights and maintaining the integrity of judicial processes. The article concludes by outlining policy recommendations and best practices for lawmakers, judiciary, and law enforcement agencies. These recommendations aim to foster responsible innovation in criminal justice AI, ensuring that technological advancements align with ethical standards and constitutional protections. Our research contributes to the ongoing dialogue on AI ethics in the legal domain and provides a roadmap for the ethical integration of AI in criminal justice systems worldwide.

Guiding the way: a comprehensive examination of AI guidelines in global media

AbstractWith the increasing adoption of artificial intelligence (AI) technologies in the news industry, media organizations have begun publishing guidelines that aim to promote the responsible, ethical, and unbiased implementation of AI-based technologies. These guidelines are expected to serve journalists and media workers by establishing best practices and a framework that helps them navigate ever-evolving AI tools. Drawing on institutional theory and digital inequality concepts, this study analyzes 37 AI guidelines for media purposes in 17 countries. Our analysis reveals key thematic areas, such as transparency, accountability, fairness, privacy, and the preservation of journalistic values. Results highlight shared principles and best practices that emerge from these guidelines, including the importance of human oversight, explainability of AI systems, disclosure of automated content, and protection of user data. However, the geographical distribution of these guidelines, highlighting the dominance of Western nations, particularly North America and Europe, can further ongoing concerns about power asymmetries in AI adoption and consequently isomorphism outside these regions. Our results may serve as a resource for news organizations, policymakers, and stakeholders looking to navigate the complex AI development toward creating a more inclusive and equitable digital future for the media industry worldwide.

Adoption of ChatGPT and other Artificial Intelligence among Physicians around the world: Barriers and the way-forward

Artificial Intelligence (AI) has emerged as a transformative force in modern society, revolutionizing various industries, including healthcare. ChatGPT, a language processing tool, has garnered significant attention for its ability to simulate human-like conversations and assist in a myriad of tasks. Despite its potential, the adoption of ChatGPT among physicians worldwide remains limited, particularly in regions like Asia, Latin America and Sub-Saharan Africa. This reluctance poses barriers to leveraging AI for novel diagnosis and treatment approaches, hindering progress in global healthcare. In this study, we explore the medical importance of AI, particularly ChatGPT, highlighting its potential to reduce medical errors and provide accurate healthcare information. However, concerns regarding bias, dependency, and plagiarism have surfaced, necessitating cautious integration and human oversight in its application. Recommendations include increasing awareness among physicians, implementing ethical guidelines, and utilizing AI-detecting software to mitigate risks. Furthermore, collaboration between international medical associations, policymakers, and governing bodies is essential to promoting AI adoption and ensuring its ethical use. While AI holds promise for advancing healthcare, its full potential hinges on overcoming current barriers and fostering a symbiotic relationship between technology and human expertise. Future research must address lingering uncertainties and ethical considerations to realize the transformative impact of AI in medicine.

Do Humans Trust AI in HRM? Why Do? Why Not? — Insights from a Decade of Research

This literature review investigates the complex issue of trust in the context of Artificial Intelligence (AI) applications in Human Resource Management (HRM). Adopting the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we systematically reviewed 43 research articles published between 2013 and 2023 to examine the attitudes towards AI as a disruptive technology in HRM practices, the ethical and legal challenges that influence trust, and the strategies for building a trustworthy future. The findings reveal a mixed picture of trust, with growing adoption of AI in various HRM practices, such as recruitment, selection, performance management, and employee development, accompanied by significant concerns related to bias, transparency, privacy, and job displacement. The review identifies key factors that affect trust in AI, including perceived usefulness, ease of use, and fairness, as well as the ethical and legal challenges that erode trust, such as data privacy, algorithmic bias, and lack of explainability. Furthermore, it discusses the implications for HRM professionals and proposes strategies for enhancing trust, such as ensuring human oversight, promoting transparency, and developing AI literacy skills. This review contributes to the research on AI in HRM by providing a nuanced understanding of the trust dynamics, challenges, and opportunities associated with this disruptive technology, offering valuable insights for both researchers and practitioners.

Leveraging Large Language Models for Direct Interaction with SysML v2

AbstractThis paper examines the potential integration of Large Language Models (LLMs) with the Systems Modeling Language version 2 (SysML v2), proposing a novel methodology for systems engineering by capitalizing on the enhanced readability and human‐friendly syntax of SysML v2. Given the emergent sophistication of LLMs and the coincidental development of SysML v2—‐an endeavor that presents a pivot toward naturally articulated model interaction—‐we explore the possibilities and implications of such an intersection. Our investigation posits that LLMs can serve not only as an interpretive layer, allowing for the syntactically simplified manipulation of system models, but also as a catalyst for a knowledge‐driven design approach.We highlight the efficiencies gained by deploying LLMs for SysML v2 interactions, which reduce the dependency on technical expertise traditionally needed for API navigation and model management. Through case studies and analysis, we demonstrate that the conversational engagement with system models facilitated by LLMs can lead to a democratized and accelerated design process. However, this advent is tempered by a critical awareness of potential pitfalls, such as automation bias and overreliance on automated systems—‐underscoring the need for continued human oversight and the examination of ethical considerations.Emphasizing the chance of SysML v2 being inherently English‐based and the parallel maturation of LLMs, this paper suggests that the collaborative utilization of these concurrent advancements may offer an opportune fusion, potentially revolutionizing the way systems are modeled and managed. Future work involves the empirical validation of these approaches and a deeper investigation into interoperability with existing and future systems engineering ecosystems. The ultimate goal is to ensure that this fusion not only complements human expertise but also propels systems engineering into a new era of innovation and holistic design.

Challenges and Prospects for the Law of Armed Conflict — A Humanitarian Perspective

The law of armed conflict, as the international legal system that regulates the conduct of war and mitigates its catastrophic consequences, has in recent years faced the challenges of high-tech forms of warfare, the involvement of non-state actors, and the complexities of civil-military integration. To address the complexity of non-international armed conflict, international humanitarian law should be “strategically” disseminated, non-government parties should be urged to issue unilateral declarations or codes of conduct, and the UN Security Council Action Plan model and the Geneva Call’s “Deed of Commitment” model should continue to be explored and implemented. The distinction between combatants and civilians in cyber-armed conflict requires more attention to actual behavior than to the appearance of the operator, with full consideration of the relationship between full membership and function. For the use of autonomous weapons systems, we must balance military interests and the consequences of damage, maximize the recognition and decision-making capabilities of autonomous weapons systems, and maintain human oversight and control during high-risk decisions.

Generative Pre-trained Transformer for Pediatric Stroke Research: A Pilot Study

BackgroundPediatric stroke is an important cause of morbidity in children. Although research can be challenging, large amounts of data have been captured through collaborative efforts in the International Pediatric Stroke Study (IPSS). This study explores the use of an advanced artificial intelligence program, the Generative Pre-trained Transformer (GPT), to enter pediatric stroke data into the IPSS. MethodsThe most recent 50 clinical notes of patients with ischemic stroke or cerebral venous sinus thrombosis at the UTHealth Pediatric Stroke Clinic were deidentified. Domain-specific prompts were engineered for an offline artificial intelligence program (GPT) to answer IPSS questions. Responses from GPT were compared with the human rater. Percent agreement was assessed across 50 patients for each of the 114 queries developed from the IPSS database outcome questionnaire. ResultsGPT demonstrated strong performance on several questions but showed variability overall. In its early iterations it was able to match human judgment occasionally with an accuracy score of 1.00 (n = 20, 17.5%), but it scored as low as 0.26 in some patients. Prompts were adjusted in four subsequent iterations to increase accuracy. In its fourth iteration, agreement was 93.6%, with a maximum agreement of 100% and minimum of 62%. Of 2400 individual items assessed, our model entered 2247 (93.6%) correctly and 153 (6.4%) incorrectly. ConclusionsAlthough our tailored generative model with domain-specific prompt engineering and ontological guidance shows promise for research applications, further refinement is needed to enhance its accuracy. It cannot enter data entirely independently, but it can be employed in tandem with human oversight contributing to a collaborative approach that reduces overall effort.

Effectiveness of Artificial Intelligence Usage as Translation Medium among English Literature Student of UIN Sunan Gunung Djati Bandung

The integration of Artificial Intelligence (AI) in translation services has transformed the landscape of linguistic studies. This study assesses the effectiveness of AI-driven translation tools among English Literature students at UIN Sunan Gunung Djati Bandung, focusing on translation accuracy, usability, and impact on language learning. Using a mixed-methods approach, the research collected quantitative data through questionnaires and supported it with a literature review. Findings indicate that 96.6% of students regularly use AI translation tools, with a majority finding them helpful. However, concerns about over-reliance on these tools and their potential negative impact on language skill development were noted. The study highlights the dual role of AI tools in enhancing learning while requiring human oversight for accuracy and contextual relevance. Recommendations for improving AI tools include ensuring contextually accurate translations, enhancing vocabulary databases, and integrating human input. Overall, while AI translation tools significantly aid students' academic activities, a balanced approach is crucial to avoid dependency and ensure the development of independent language skills

AI liability in Europe: How does it complement risk regulation and deal with the problem of human oversight?

Who should compensate you if you get hit by a car in “autopilot” mode: the safety driver or the car manufacturer? What about if you find out you were unfairly discriminated against by an AI decision-making tool that was being supervised by an HR professional? Should the developer compensate you, the company that procured the software, or the (employer of the) HR professional that was “supervising” the system's output?These questions do not have easy answers. In the European Union and elsewhere around the world, AI governance is turning towards risk regulation. Risk regulation alone is, however, rarely optimal. The situations above all involve the liability for harms that are caused by or with an AI system. While risk regulations like the AI Act regulate some aspects of these human and machine interactions, they do not offer those impacted by AI systems any rights and little avenues to seek redress. From a corrective justice perspective risk regulation must also be complemented by liability law because when harms do occur, harmed individuals should be compensated. From a risk-prevention perspective, risk regulation may still fall short of creating optimal incentives for all parties to take precautions.Because risk regulation is not enough, scholars and regulators around the world have highlighted that AI regulations should be complemented by liability rules to address AI harms when they occur. Using a law and economics framework this Article examines how the recently proposed AI liability regime in the EU – a revision of the Product Liability Directive, and an AI Liability effectively complement the AI Act and how they address the particularities of AI-human interactions.

Real Estate Insights: Establishing transparency – setting AI standards in property valuation

PurposeThis real estate insight scrutinises the emerging role of Artificial Intelligence (AI), particularly Large Language Models (LLMs) like ChatGPT, in property valuation, advocating for establishing standardised reporting guidelines in AI-enabled property valuation.Design/methodology/approachThrough a conceptual exploration, this piece examines the shift towards AI integration in property valuation and the critical role of Explainable Artificial Intelligence (XAI) in this transition. It discusses the CANGARU framework for developing inclusive and universally applicable reporting guidelines and the importance of human oversight in validating AI-enabled valuations.FindingsIntegrating LLMs into property valuation signifies potential efficiency gains and task automation but also introduces risks related to accuracy, bias, and ethical dilemmas. Standardised reporting guidelines are identified as essential for responsibly harnessing AI’s benefits.Practical implicationsThe article underscores the need for the real estate industry to adopt transparent reporting practices, with valuers acting as expert interpreters of AI outputs. Emphasising error reporting in XAI not only aids in understanding AI-generated insights but also builds trust among stakeholders, ensuring AI’s ethical and effective application in property valuation.Originality/valueThis commentary contributes to the discourse on AI’s role in property valuation by focusing on the need for standard reporting guidelines that align with professional standards and legal frameworks. It advocates for a balanced approach to AI integration, where technological advancements complement traditional valuation expertise, ensuring accurate, fair, and transparent property valuations.

Advanced AI Solutions for Securities Trading: Building Scalable and Optimized Systems for Global Financial Markets

Integrating Artificial Intelligence (AI) into securities trading software has revolutionized the financial markets by enhancing scalability, efficiency, and optimization. This paper explores the historical evolution and advancements in AI-driven trading systems, emphasizing their impact on global financial markets. The study investigates how machine learning, deep learning, and other AI technologies enable sophisticated trading strategies, improve market liquidity, and reduce transaction costs. It also addresses the challenges AI integration poses, including decision-making opacity, bias, increased market volatility, and data privacy concerns. The paper argues for the need for scalable architectures, optimization algorithms, and transparent governance to mitigate these issues. Additionally, human oversight remains crucial for evaluating AI outputs and maintaining accountability. This research aims to provide a comprehensive analysis of advanced AI systems in securities trading, highlighting their potential to enhance market efficiency and stability.

Automation bias in public administration – an interdisciplinary perspective from law and psychology

The objective of this paper is to break down the widely presumed dichotomy, especially in law, between fully automated decisions and human decisions from a psychological and normative perspective. This is particularly relevant as human oversight is seen as an effective means of quality control, including in the current AI Act. The phenomenon of automation bias argues against this assumption. We have investigated this phenomenon of automation bias, as a behavioral effect of and its implications in normative institutional decision-making situations. The phenomenon of automation bias, whereby individuals overly rely on machine-generated decisions or proposals, has far-reaching implications. Excessive reliance may result in a failure to meaningfully engage with the decision at hand, resulting in an inability to detect automation failures, and an overall deterioration in decision quality, potentially up to a net-negative impact of the decision support system. As legal systems emphasize the role of human decisions in ensuring fairness and quality, this paper critically examines the inadequacies of current EU and national legal frameworks in addressing the risks of automation bias. Contributing a novel perspective, this article integrates psychological, technical, and normative elements to analyze automation bias and its legal implications. Anchoring human decisions within legal principles, it navigates the intersections between AI and human-machine interactions from a normative point of view. An exploration of the AI Act sheds light on potential avenues for improvement. In conclusion, our paper proposes four steps aimed at effectively countering the potential perils posed by automation bias. By linking psychological insights, legal analysis, and technical implications, this paper advocates a holistic approach to evolving legal frameworks in an increasingly automated world.

Trust in automation and the accuracy of human–algorithm teams performing one-to-one face matching tasks

The human face is commonly used for identity verification. While this task was once exclusively performed by humans, technological advancements have seen automated facial recognition systems (AFRS) integrated into many identification scenarios. Although many state-of-the-art AFRS are exceptionally accurate, they often require human oversight or involvement, such that a human operator actions the final decision. Previously, we have shown that on average, humans assisted by a simulated AFRS (sAFRS) failed to reach the level of accuracy achieved by the same sAFRS alone, due to overturning the system’s correct decisions and/or failing to correct sAFRS errors. The aim of the current study was to investigate whether participants’ trust in automation was related to their performance on a one-to-one face matching task when assisted by a sAFRS. Participants (n = 160) completed a standard face matching task in two phases: an unassisted baseline phase, and an assisted phase where they were shown the identification decision (95% accurate) made by a sAFRS prior to submitting their own decision. While most participants improved with sAFRS assistance, those with greater relative trust in automation achieved larger gains in performance. However, the average aided performance of participants still failed to reach that of the sAFRS alone, regardless of trust status. Nonetheless, further analysis revealed a small sample of participants who achieved 100% accuracy when aided by the sAFRS. Our results speak to the importance of considering individual differences when selecting employees for roles requiring human–algorithm interaction, including identity verification tasks that incorporate facial recognition technologies.

"How I would like AI used for my imaging": children and young persons' perspectives.

Artificial intelligence (AI) tools are becoming more available in modern healthcare, particularly in radiology, although less attention has been paid to applications for children and young people. In the development of these, it is critical their views are heard. A national, online survey was publicised to UK schools, universities and charity partners encouraging any child or young adult to participate. The survey was "live" for one year (June 2022 to 2023). Questions about views of AI in general, and in specific circumstances (e.g. bone fractures) were asked. One hundred and seventy-one eligible responses were received, with a mean age of 19 years (6-23 years) with representation across all 4 UK nations. Most respondents agreed or strongly agreed they wanted to know the accuracy of an AI tool that was being used (122/171, 71.3%), that accuracy was more important than speed (113/171, 66.1%), and that AI should be used with human oversight (110/171, 64.3%). Many respondents (73/171, 42.7%) felt AI would be more accurate at finding problems on bone X-rays than humans, with almost all respondents who had sustained a missed fracture strongly agreeing with that sentiment (12/14, 85.7%). Children and young people in our survey had positive views regarding AI, and felt it should be integrated into modern healthcare, but expressed a preference for a "medical professional in the loop" and accuracy of findings over speed. Key themes regarding information on AI performance and governance were raised and should be considered prior to future AI implementation for paediatric healthcare. Artificial intelligence (AI) integration into clinical practice must consider all stakeholders, especially paediatric patients who have largely been ignored. Children and young people favour AI involvement with human oversight, seek assurances for safety, accuracy, and clear accountability in case of failures. Paediatric patient's needs and voices are often overlooked in AI tool design and deployment. Children and young people approved of AI, if paired with human oversight and reliability. Children and young people are stakeholders for developing and deploying AI tools in paediatrics.

Assessing the Decision-Making Capabilities of Artificial Intelligence Platforms as Institutional Review Board Members.

Background: Institutional review boards (IRBs) face delays in reviewing research proposals, underscoring the need for optimized standard operating procedures (SOPs). This study assesses the abilities of three artificial intelligence (AI) platforms to address IRB challenges and draft essential SOPs. Methods: An observational study was conducted using three AI platforms in 10 case studies reflecting IRB functions, focusing on creating SOPs. The accuracy of the AI outputs was assessed against good clinical practice (GCP) guidelines. Results: The AI tools identified GCP issues, offered guidance on GCP violations, detected conflicts of interest and SOP deficiencies, recognized vulnerable populations, and suggested expedited review criteria. They also drafted SOPs with some differences. Conclusion: AI platforms could aid IRB decision-making and improve review efficiency. However, human oversight remains critical for ensuring the accuracy of AI-generated solutions.

Bias and ethics of AI systems applied in auditing - A systematic review

The integration of artificial intelligence into auditing shows great potential in enhancing automation and gaining insights from complex data. However, it also presents significant ethical challenges, including algorithmic biases, transparency, accountability, and fairness. This study aimed to investigate the sources of bias and risks posed by AI systems applied in auditing and the complex downstream interactions and effects they have. The study also explored the technical and ethical guardrails proposed and recommendations for translating principles into auditing practice. A systematic methodology was employed to acquire relevant studies across scientific databases. This involved a three-step process, including a targeted search query using Boolean operators and snowballing to yield 310 preliminary publications. A systematic review process was then conducted to identify 123 relevant articles focused on AI's implications for auditing, accounting, finance, or assurance contexts. Finally, screening and filtering on research quality distilled 83 high-quality publications from the year 2018 to 2023 spanning computer science, accounting, management science, and ethics disciplines. The analysis revealed five primary sources driving technical and human biases: data deficiencies, demographic homogeneity, spurious correlations, improper comparators, and cognitive biases. It also highlighted wider issues, such as trade-offs between efficiency and diligence, erosion of human skills and judgement, data dependence risks, and privacy violations from uncontrolled personal data exploitation. The study found promising remedies, including causal modeling to enable auditors to uncover subtle biases, representative algorithmic testing to evaluate fairness, periodic auditing of AI systems, human oversight alongside automation, and embedding ethical values like fairness and accountability into system design. The study concludes that auditors play a crucial role in assessing and ensuring AI's reliable and socially beneficial integration. It recommends governance, risk assessment before deployment, ongoing performance monitoring, and policies fostering trust and collaboration to responsibly translate principles into auditing practice.

Docimological Quality Analysis of LLM-Generated Multiple Choice Questions in Computer Science and Medicine

Assessment is an essential part of education, both for teachers who assess their students as well as learners who may evaluate themselves. Multiple-choice questions (MCQ) are one of the most popular types of knowledge assessment, e.g., in medical education, as they can be automatically graded and can cover a wide range of learning items. However, the creation of high-quality MCQ items is a time-consuming task. The recent advent of Large Language Models (LLM), such as Generative Pre-trained Transformer (GPT), caused a new momentum for automatic question generation solutions. Still, evaluating generated questions according to the best practices for MCQ item writing is needed to ensure docimological quality. In this article, we propose an analysis of the quality of LLM-generated MCQs. We employ zero-shot approaches in two domains, namely computer science and medicine. In the former, we make use of 3 GPT-based services to generate MCQs. In the latter, we developed a plugin for the Moodle learning management system that generates MCQs based on learning material. We compare the generated MCQs against common multiple-choice item writing guidelines. Among the major challenges, we determined that while LLMs are certainly useful in generating MCQs more efficiently, they sometimes create broad items with ambiguous keys or implausible distractors. Human oversight is also necessary to ensure instructional alignment between generated items and course contents. Finally, we propose solutions for AQG developers.