Don’t give it away: the hidden risks of generative AI for scientists

Introduction: This paper discusses the ethical and adversarial implications of implementing generative AI in military cyber secure methods. Generative AI has been displayed in numerous applications for threat simulation and defense from threats in civilian use. Still, there are important ethical considerations in military use because of the potential misuse of generative AI. Cyber threats against military systems continue to grow more sophisticated than previously, and we hope to add data to the body of research in this area to help bridge the identified gap in understanding the risks of generative AI in a military context. Objectives: The paper seeks to explore the ethical dilemmas, including accountability, autonomy, and misuse, surrounding military applications of generative AI. The paper examines adversarial risks associated with generative AI, including manipulation or other uses by hostile actors. The objective is to recommend measures for considering the ethical dilemmas, while at the same time improving the defenses. Methods: The methodology will assess ethical risks such as autonomy, weaponization, and bias related to AI systems. It will determine adversarial risks by recommending using adversarial training strategies, hybrid AI systems, and robust defense mechanisms against adversarially manipulated AI-generated threats. It will also propose ethical frameworks and accountability models for military cybersecurity. Results: This paper provides a comparative performance evaluation of military cybersecurity systems in a traditional and an AI-smart cyber context. The significant findings establish that generative AI potentially improves detection accuracy and, most notably, response times. It also introduces new risks such as adversarial manipulation. The experimentation results illustrated how adversarial training increases the robustness of models, reduces vulnerability, and provides greater defensive capabilities against adversarial threats. Conclusions: Generative AI in military cybersecurity has considerable benefits compared to traditional methods, particularly in enhancing detection performance, response time, and adaptability. As illustrated, the benefits of an AI-enhanced system improved the accuracy of malware detection by 15%, from 80% to 95%, and a 15% increase in phishing email detection, from 78% to 93%. The ability to react quickly to a new threat was also key, as response time was reduced by 60%, from 5 minutes to 2 minutes, which is essential in military situations where responding quickly will minimize impact. Additionally, the AI systems showed the ability to reduce false favorable rates from 10% to 4% (which is excellent) and lower false negative rates from 12% to 5% (which is also that employed the AI system greatly based on its ability to identify what a real threat looks like and its=apability to identify a real threat.

As generative AI technologies advance, understanding their capability to emulate human-like experiences in marketing communication becomes crucial. This research examines whether generative AI can create experiential narratives that resonate with humans in terms of embodied cognition, affect, and lexical diversity. An automated text analysis reveals that while reviews generated by ChatGPT 3.5 exhibit lower levels of embodied cognition and lexical diversity compared with reviews by human experts, they display more positive affect (Study 1a). However, human raters struggle to notice these differences, rating half of the selected reviews from AI higher in embodied cognition and usefulness (Study 1b). Instances of hallucination in AI-generated content were detected by human raters. For social media posts, the more sophisticated ChatGPT 4 model demonstrates superior perceived lexical diversity and leads to higher purchase intentions in unbranded content compared with human copywriters (Study 2). This research evaluates the performance of large language models in generating experiential marketing narratives. The comparative studies reveal the models’ strengths in presenting positive emotions and influencing purchase intent while identifying limitations in embodied cognition and lexical diversity compared with human-authored content. The findings have implications for marketers and policy makers in understanding generative AI's potential and risks in marketing.

The revolutionary way that generative AI disrupts business models today is by delivering radical shifts in important things like value delivery, business operations, and competitiveness. It allows businesses to perform difficult tasks, orchestrate customized client interactions, and foster change throughout sectors. With endless possibilities, generative AI's risks upon implementation are as follows: Data protection, Ethical and legal spheres, and Talent demand. Some challenges must be overcome when adopting AI in any organization to harness this technology for work benefits. Generative AI will continue to define future business models ' directions and contours through innovation, support of the decision-making process, and a focus on the principles of ethical entrepreneurship. This paper discusses the opportunities of generative AI for business model innovations, its effect on operational improvement, and entering a new kind of competition in a market. Finally, it reveals the factors that companies experience when implementing AI and presents the possible future trend of deploying business models based on artificial intelligence.

Objective – To examine how generative AI landscapes can function as instruments of environmental and social awareness or reinforce processes of symbolic alienation. Methodology – An interdisciplinary literature review combined with the analysis of news reports and case studies, including UrbanGenAI and environmental scenario simulations using landscape generative AIs. Originality/Relevance – Offers an unprecedented discussion of the ambivalence of digital AI landscapes, intertwining aesthetics, politics, and ecology within a contemporary debate. Results – Demonstrates that generative landscapes expand planning and awareness capabilities, yet may also foster illusions, greenwashing, and misinformation. Theoretical/Methodological Contributions – Integrates concepts from geography, philosophy, arts, and technology, proposing a critical and comparative approach to the potentials and risks of AI. Social and Environmental Contributions – Highlights the use of AI-generated images as pedagogical, political, and mobilization tools, provided they are applied ethically and responsibly.

Objective – To examine how generative AI landscapes can function as instruments of environmental and social awareness or reinforce processes of symbolic alienation. Methodology – An interdisciplinary literature review combined with the analysis of news reports and case studies, including UrbanGenAI and environmental scenario simulations using landscape generative AIs. Originality/Relevance – Offers an unprecedented discussion of the ambivalence of digital AI landscapes, intertwining aesthetics, politics, and ecology within a contemporary debate. Results – Demonstrates that generative landscapes expand planning and awareness capabilities, yet may also foster illusions, greenwashing, and misinformation. Theoretical/Methodological Contributions – Integrates concepts from geography, philosophy, arts, and technology, proposing a critical and comparative approach to the potentials and risks of AI. Social and Environmental Contributions – Highlights the use of AI-generated images as pedagogical, political, and mobilization tools, provided they are applied ethically and responsibly.

Current advanced AI technologies like LLM and popular generative AI technologies based on generative adversarial networks (GANs) are revolutionising industries. Yet, it brings about severe cybersecurity threats at the same time. This chapter goes over one general risk of generative AI: fake information, or Generative AI Misinformation and Deepfakes, and six more specific risks: Generative AI Phishing, Generative AI Data Poisoning, Malicious Code Generation by Generative AI, and Privacy Violation by Generative AI. It raises the purpose of misuse with an elaboration of social engineering, smart malware, and leakage of sensitive data. This chapter provides a framework of measures: using detection tools, fortifying defences against phishing attacks, securing training data, proper control of AI misuse, and better data privacy regulation. Minimizing these risks, organizations, and stakeholders will be able to safely take advantage of reasons for the implementation of generative AI.

Every major technical invention resurfaces the dual-use dilemma—the new technology has the potential to be used for good as well as for harm. Generative AI (GenAI) techniques, such as large language models (LLMs) and diffusion models, have shown remarkable capabilities (e.g., in-context learning, code-completion, and text-to-image generation and editing). However, GenAI can be used just as well by attackers to generate new attacks and increase the velocity and efficacy of existing attacks. This monograph reports the findings of a workshop held at Google (co-organized by Stanford University and the University of Wisconsin-Madison) on the dual-use dilemma posed by GenAI. This work is not meant to be comprehensive, but is rather an attempt to synthesize some of the interesting findings from the workshop. We discuss short-term and long-term goals for the community on this topic. We hope this work provides both a launching point for a discussion on this important topic as well as interesting problems that the research community can work to address.

Drawing on the Theory of Planned Behaviour (TPB), this study investigates the relationship between the perceived benefits, strengths, weaknesses, and risks of generative AI (GenAI) tools and the fundamental factors of the TPB model (i.e., attitude, subjective norms, and perceived behavioural control). The study also investigates the structural association between the TPB variables and intention to use GenAI tools, and how the latter might affect the actual usage of GenAI tools in higher education. The paper adopts a quantitative approach, relying on an anonymous self-administered online questionnaire to gather primary data from 130 lecturers and 168 students in higher education institutions (HEIs) in several countries, and PLS-SEM for data analysis. The results indicate that although lecturers' and students' perceptions of the risks and weaknesses of GenAI tools differ, the perceived strengths and advantages of GenAI technologies have a significant and positive impact on their attitudes, subjective norms, and perceived behavioural control. The TPB core variables positively and significantly impact lecturers' and students’ intentions to use GenAI tools, which in turn significantly and positively impact their adoption of such tools. This paper advances theory by outlining the factors shaping the adoption of GenAI technologies in HEIs. It provides stakeholders with a variety of managerial and policy implications for how to formulate suitable rules and regulations to utilise the advantages of these tools while mitigating the impacts of their disadvantages. Limitations and future research opportunities are also outlined.

Generative AI tools stand at the threshold of innovation and the erosion of the long-standing values of creativity, critical thinking, authorship, and research in higher education. This research crafted a novel framework from the technology, organization, and environment (TOE) framework to guide higher educational institutions in Nigeria to navigate the ethical dilemma of generative AI. A questionnaire was used to collect data from twelve higher institutions among lecturers, students, and researchers across the six (6) geopolitical zones of Nigeria. The structural equation modeling was used to analyze the data using the SPPS Amos version 23. The results revealed that factors such as perceived risks of generative AI, Curriculum support, institutional policy, and perceived generative AI trends positively impact the need for a generative AI ethical framework in higher educational institutions in Nigeria. Furthermore, the study contributes to the adoption of theory to navigate the ethical dilemma in the use of generative AI tools in higher educational institutions in Nigeria. It also provides some practical implications that suggest the importance of inculcating ethical discussions into the curriculum as part of institutional policy to create awareness and guidance on the use of generative AI.

Generative AI is regarded as a major disruption to software development. Platforms, repositories, clouds, and the automation of tools and processes have been proven to improve productivity, cost, and quality. Generative AI, with its rapidly expanding capabilities, is a major step forward in this field. As a new key enabling technology, it can be used for many purposes, from creative dimensions to replacing repetitive and manual tasks. The number of opportunities increases with the capabilities of large-language models (LLMs). This has raised concerns about ethics, education, regulation, intellectual property, and even criminal activities. We analyzed the potential of generative AI and LLM technologies for future software development paths. We propose four primary scenarios, model trajectories for transitions between them, and reflect against relevant software development operations. The motivation for this research is clear: the software development industry needs new tools to understand the potential, limitations, and risks of generative AI, as well as guidelines for using it.

Generative AI is constantly developing with its unique competitiveness, and its content influence is gradually penetrating into people's daily life, but at the same time, it has also caused a series of discussions on the regulation of artificial intelligence legal risks. This paper analyzes and researches ChatGPT and finds many existing problems, such as data risk, algorithm model training risk, and content generation risk. In order to better deal with the risks of generative AI, it is necessary to find a healthy path conducive to the development of generative AI from the perspectives of database establishment, computing legal system, data compliance construction, scientific and technological ethics governance, and improvement of laws and regulations.

This research explores the ethical considerations of using generative AI in application development. With the rapid adoption of tools like ChatGPT, GitHub Copilot, and DALL- E, generative AI has shown significant potential to enhance developer productivity and creativity. However, these technologies also raise ethical concerns. Which is including inherent biases, data privacy risks, and the potential negative impact on developer skills and creativity. This report examines the strengths and weaknesses of generative AI, its implications on software development, and provides recommendations for responsible adoption. Emphasizing a human-in-the-loop approach, transparency, and adherence to ethical guidelines, we aim to balance the benefits and risks of generative AI to support ethical application development.

The rise of Generative Artificial Intelligence (Generative AI) offers transformative potential for productivity and creativity but also introduces significant risks that challenge current AI regulation frameworks. This study systematically investigates these risks, including security vulnerabilities, privacy concerns, copyright infringements, and algorithmic biases. It critically assesses the effectiveness of existing regulatory approaches in managing these issues. Through a comparative analysis of regulatory practices in the European Union, the United States, the UK, and China, the study reveals diverse strategies ranging from stringent risk-based models to flexible market-driven approaches. The findings underscore the need for a dynamic and adaptable regulatory framework that can effectively balance the rapid advancement of Generative AI with the imperative to protect public interest and promote innovation. This paper concludes by advocating for the development of adapting regulatory approaches to address the evolving challenges posed by Generative AI.

Against the backdrop of accelerating digital transformation, GPT-based generative AI technologies are gradually penetrating the entire corporate financial disclosure process, exerting a significant dual impact on disclosure quality. Drawing on information asymmetry theory and principal-agent theory, combined with KPMG's global research data and case studies such as Amazon and AllHere, this paper systematically analyzes the positive impact and potential risks of generative AI on the quality of financial disclosure. The study finds that generative AI can reduce disclosure redundancy through automated processing, compressing MD&A report summaries to 25% of the original while retaining core information, while also improving forecast accuracy and compliance efficiency. However, this also presents risks such as "AI whitewashing," data fabrication, and algorithmic black box manipulation. For example, the US AI startup AllHere overstated its revenue by nearly 700 times by fabricating AI-related financial data. The study further suggests the need to establish a coordinated mechanism across three dimensions: optimizing corporate governance, upgrading regulatory technology, and managing model security. The conclusions indicate that the impact of generative AI on disclosure quality is not one-way; its ultimate effect depends on the alignment between technical application specifications and risk prevention and control systems. This finding provides empirical evidence for companies to rationally utilize AI technology and for regulators to improve governance rules.

This study provides empirical data on the impact of generative AI in education, with special emphasis on sustainable development goals (SDGs). By conducting a thorough analysis of the relationship between generative AI technologies and educational outcomes, this research fills a critical gap in the literature. The insights offered are valuable for policymakers seeking to leverage new educational technologies to support sustainable development. Using Smart-PLS4, five hypotheses derived from the research questions were tested based on data collected from an E-Questionnaire distributed to academic faculty members and education managers. Of the 311 valid responses, the measurement model assessment confirmed the validity and reliability of the data, while the structural model assessment validated the hypotheses. The study’s findings reveal that New Approaches to Learning Outcome Assessment (NALOA) significantly contribute to achieving SDGs, with a path coefficient of 0.477 (p < 0.001). Similarly, the Use of Generative AI Technologies (UGAIT) has a notable positive impact on SDGs, with a value of 0.221 (p < 0.001). A Paradigm Shift in Education and Educational Process Organization (PSEPQ) also demonstrates a significant, though smaller, effect on SDGs with a coefficient of 0.142 (p = 0.008). However, the Opportunities and Risks of Generative AI in Education (ORGIE) study did not find statistically significant evidence of an impact on SDGs (p = 0.390). These findings highlight the potential opportunities and challenges of using generative AI technologies in education and underscore their key role in advancing sustainable development goals. The study also offers a strategic roadmap for educational institutions, particularly in Oman to harness AI technology in support of sustainable development objectives.