Policies and regulations of artificial intelligence in healthcare, finance, agriculture, manufacturing, retail, energy, and transportation industry

The rapid evolution of artificial intelligence (AI) technologies has required to standardized policies and regulations to provide proper care, safeguard, and equity in the industry. This study revisits the field to identify the AI policy and regulation frameworks that nowadays are being implemented and their most relevant issues. AI is entering - or rather, already exists in - everything from health to finance, and across the globe regulators and governments are now looking at both the need for innovation and the requirement of oversight. Some notable programmes include the European Union's AI Act, where expects to classify AI systems by 'risk levels' and have more stringent requirements for high-risk applications. In the United States, guidelines around transparency, accountability, and debiasing. This reinforces the importance of public-private partnerships (PPPs) and a shared approach in formulating agile and future-proof regulations. The up-and-coming trends can be seen in regulatory sandboxes for AI pilots for developing AI innovations in a controlled environment and AI ethics boards driving corporate practices towards AI. It also reflects on the effects these regulations might have on innovation and the dynamics of the market, suggesting that, although difficult, regulation is necessary to support public trust and secure the sustainable development of AI technologies. Keywords: Policy, Regulation, Artificial intelligence, Decision making, Decision support system, Machine learning, Deep learning Citation: Rane, N. L., Paramesha, M., Rane, J., & Mallick, S. K. (2024). Policies and regulations of artificial intelligence in healthcare, finance, agriculture, manufacturing, retail, energy, and transportation industry. In Artificial Intelligence and Industry in Society 5.0 (pp. 67-81). Deep Science Publishing. https://doi.org/10.70593/978-81-981271-1-2_4 4.1 Introduction The runaway development pace of artificial intelligence (AI) technologies has provided impetus for profound changes across different industries; such a roll-out is naturally accompanied by co-evolution of the policy and regulatory landscape (Wischmeyer, & Rademacher, 2020; Hoffmann-Riem, 2020; de Almeida et al., 2021). As AI integrates into the health, financial, production, and transportation sectors, amongst others, so is the establishment of regulatory solid frameworks to oversee its deployment and reduce associated risks (Erdélyi, & Goldsmith, 2018; Lauterbach, 2019; Taeihagh, 2021;). The twin challenge for policymakers is encouraging innovation in the early stages while ensuring that AI systems meet ethical standards for safety and engender public trust (Wischmeyer, & Rademacher, 2020; Hoffmann-Riem, 2020; Paramesha et al., 2024a). Balancing creative responses to both challenges, therefore, really needs to be underpinned by a good grasp of AI's technological possibilities as well as its socio-economic effects from deployment. Policy-wise, regulatory environments around AI have been historically reactive, for example, setting standards in reaction to a specific issue rather than as proactive standard-setting. These approaches have led to a fragmented regulatory landscape, showing high inconsistency across jurisdictions and sectors. International organizations and national governments have recently undertaken different initiatives to clarify AI policies, thus recognizing the urgency of harmonized and forward-looking regulatory frameworks in this area. These frameworks should address central issues, such as data privacy, algorithm transparency, accountability, and how AI may deepen pre-existing inequalities (Manheim, & Kaplan, 2019; Capraro, et al., 2024; Rane et al., 2024a). Specifically, research on the policy and regulation of AI within the academic community has quickened its pace, engendering an enormous body of literature across legal studies, ethics, economics, and technology (Cath, 2018; Wong, 2021; Paramesha et al., 2024b; Rane et al., 2024b). The impact of different kinds of regulation on AI innovation and how practical various policy approaches have been probed using a raft of methodologies by researchers. This research contributes to the continuous discourse by conducting a careful literature review on AI policy and regulation in industry. Contributions of the present study: This study provides a comprehensive synthesis of literature available on AI policy and its regulation, with themes of broad consensus and divergence. This study uses sophisticated text-mining techniques to identify popular topics under discussion and their interrelations, attaining a granulated understanding of the current discourse. This study uses statistical methods to find out different clusters of related studies, pointing out emerging trends likely to receive significant attention in the future. 4.2 Methodology This research approach policy and regulation issues relevant to AI in the industry with a comprehensive review of the literature. The identification of academic articles, policy papers, and industry reports is done through databases such as Google Scholar, Scopus, and Web of Science. The literature search focused on thorough literature gathering guided by keywords like "artificial intelligence," "AI policy," "AI regulation," "industrial AI," and "AI governance.". Literature data is then fed into bibliometric software, VOSviewer, for co-occurrence analysis to establish how often the keywords appeared and how much they relate. By cluster analysis of the co-occurrence network, it is possible to further divide this literature into distinct groups regarding thematic similarities. Each cluster represents a particular aspect of AI policy and regulation and allows detailed examinations of subtopics: ethical issues, regulatory frameworks, and industry-specific challenges. This methodological approach gives a systematic and structured review of the existing body of knowledge, helping to deepen an understanding of the complex nature of AI governance in industry. 4.3 Results and discussions Co-occurrence and cluster analysis of the keywords The broader theme of the network also demonstrates an idea that "artificial intelligence" is vital to the frontier of research. Here, various nodes semantically similar with the central node are connected as shown in the Fig. 4.1, showing the applications of AI, which explains the interdisciplinary nature of AI. This is an important cluster in the network that deals with "decision making" and "decision support system". This family is tightly integrated with AI, means a bundle of important concepts are covered within this family like machine learning, deep learning, neural networks and reinforcement learning. These connections illustrate why AI is critical to the decision-making processes in a wide array of industries. When machine learning and deep learning algorithms are integrated into decision support systems, then the accuracy of predictions and the efficiency of problem-solving are increased by manifolds. Another large group relates to ethics, privacy, and regulation - which is significant. This cluster demonstrates the increasing attention to ethics in AI, data privacy, and regulation. The relationship with these terms and AI demonstrates the dialogue and inquiry about making AI technologies more robust which continues through the age. Fig. 4.1 Co-occurrence analysis of the keywords in literature The final execution with the sense of utmost responsibility. This cluster is about how governance shapes the social implications of AI. Example, the existence of the words "public policy", "policy making," and "laws and legislation" is clear. The diagram also connects "sustainable development" and "energy policy. Themes link this cluster to artificial intelligence, including energy efficiency, energy consumption, and sustainability. The connections demonstrated here are cases of how AI can make a real impact in this kind of sustainability; by helping to save energy and promote green policy. The powerful partnership of AI in energy management systems will go a long way towards addressing sustainability objectives. Another captivating bundle of topics belong to the intersection of "healthcare policy", "public health" and "covid-19." Terms that appear in this cluster are related to health care delivery, information processing and clinical knowledge, which is relevant to artificial intelligence. This widespread application of such terminologies is representative of how prominently AI has come to the forefront in the health care sector - more so in the era of the Covid-19 pandemic. AI has helped with advancements in diagnostic tools, more streamlined management of health data and has provided invaluable help to public health projects. The network diagram shows in addition to a cluster of genetics, gene expression, and computational biology. The aim of this cluster is to focus on the AI application in biological and medical research. The relationships among these key words and "artificial intelligence" reinforce the role of AI both in elucidating genetic information and its regulation. AI-based algorithms help to better understand complex biological data and are in the vicinity of communities such as genomics and personalized medicine. In addition, the presence of the terms "internet," "social media," and "students" to the network hints at the rapid encroaching of AI into the arenas of digital technology and education. The onset of AI technologies in these fields has begun to change the way information is stored and shared; and affecting most areas in the society. The network diagram reveals the significance of cross-disciplinary research in the realm of AI. The interactions of fields like ethics, healthcare, energy policy, computational biology with each other clearly explain AI in its broadest contexts as well as the practical applications of them. A single-discipline approach will not work for the multi-faceted problems that AI presents. It is important to take the interdisciplinary approach, which our societies need to adopt anyway. Policy and regulations of artificial intelligence in industry Global regulatory landscape The regulatory landscapes of AI differ from country to country and region to region because they entail very different legal, cultural, and economic environments. The AI Act proposed, establish a broad framework for regulating AI within the European Union. In this sense, the legislation will categorize AI applications in terms of risk as minimal, limited, or high and apply more stringent requirements on the high-risk ones. It also means stringent testing, documentation, and transparency requirements to reduce the possible harms. In the United States case, AI regulation is more sector-specific and less centralized. Several federal agencies, such as the Food and Drug Administration (FDA) and the Federal Trade Commission (FTC), are responsible for AI applications in their domains. The National Institute of Standards and Technology (NIST) is also developing a voluntary framework that will guide the development and deployment of AI, outlining principles like transparency, fairness, and accountability. However, it is the more hands-on approach that China has pursued in AI policymaking. The concept is fully embodied in China's New Generation Artificial Intelligence Development Plan, which sets ambitious targets for AI leadership by 2030, coupled with regulations that manage data security, algorithmic transparency, and ethical standards. Ethical and safety considerations AI policy-making necessarily fares with ethical concerns (Vesnic-Alujevic et al., 2020; de Almeida et al., 2021; Paramesha et al., 2024c; Rane et al., 2024c). The ability of AI to further already existing or biased situations has put increased scrutiny on algorithmic fairness. Regulations now often require developers to implement the possibility of detecting and mitigating bias in AI systems. Under the new Act on AI, for example, periodic evaluations and activity documentation are required to ensure conformance to ethical standards by the EU. Another significant concern is safety, for applications such as and necessary stringent and of AI systems to under the most and The in its related to AI-based medical that a need for using where for be with and AI systems often require of data privacy, and have of AI regulation 2018; et al., Rane et al., The European has high standards for data which how AI systems are to and data under the with data management impact and for It is the that similar for data through the Act and an impact on AI. These their from to and out of and transparency AI and transparency are the systems will public trust Almeida et al., 2021; et al., Rane et al., The new from regulators for AI developers to clearly explain how are by their this is very critical in areas such as the of of or with regulatory frameworks AI systems that make is and 2020; 2021; Rane et al., This and of the AI and allows for and accountability. International and AI is in it for the of regulatory standards (Erdélyi, & Goldsmith, 2018; de 2021; Rane et al., like International for and the Institute of and are to AI standards. These across and ethical considerations at and The AI by various a framework for the responsible development and of AI. These principles transparency, and as principles policy should at the national while for regulations and risks AI differ across which in for different regulatory in the health sector around safety, of health and The has its regulatory for AI-based medical in which are to provide about safety and also to to the of in the the focus is on how they can and The standards on and deployment are on and to that they can in different environments. others, regulating ensuring this include the National Administration in the and the European New in AI regulation of the most critical in policy and regulation is transparency and the ability to provide et al., Taeihagh, 2021; Paramesha et al., AI systems, and using machine learning, are that how are This can be very in like health, or finance, where AI is the of to or AI trust in these systems. In that at this should establish frameworks that AI developers to make AI and on how within the AI are Another critical challenge is that of bias and in AI 2019; de Almeida et al., 2021). in most AI systems can be with that already present in society. such are not at they might further be by the AI algorithm and example, AI systems for or shown to groups than It is to regulations where and are in AI systems. This not of data and algorithmic fairness, for example, also within the that these technologies to ensure a broader AI and Another of in AI policy is related to the concerns of (Manheim, & Kaplan, 2019; et al., Paramesha et al., Rane et al., AI systems require of data to be concerns the and of data and cases have around among the need to the that data analysis has the need to This in stringent data policies like the of the European with increased data and for AI is regulatory regulatory processes often take their in find it to pace with the AI This may regulations that are or a to innovation by requirements on emerging technologies. more agile regulatory approaches that can to technological It may regulatory sandboxes within which new applications of AI be in a critical concerns in this are the economic and the of AI can through the of that are by very being to and economic they also in of and economic The will require policymakers to promote and for that will with relevant for an by AI, on the and development of social safety for the by the nature of AI development and deployment it in terms of policy and regulation. The and of AI technologies are this the need for and of the regulations of national regulatory which complex in of for and to a regarding the standards. International organizations and have to way to frameworks that ensure standards of Ethical considerations AI development and et al., 2021; et al., 2020; Rane et al., that AI systems are relevant in making and ethical regarding the of an AI have to make in situations to complex ethical and the public in guidelines to address these Another that gives to solid concerns the risks associated with AI 2020; et al., AI is to different of including data or that may its and in critical health, energy, or finance, this might have should standards and practices to AI systems from The is that of and governance of AI. on making AI more powerful and ensuring that these AI systems under and that their with will be This with distinct of accountability, to AI from 4.1 provides a structured of the various and associated with regulatory frameworks, as well as their on society. These include which concern and related to and and to broader and other more the the the increased and broader effects and other These regulations their final on economic of and The need this comprehensive analysis to the regulatory implications so that be to 4.1 of the various and associated with regulatory frameworks, as well as their on other efficiency - and - - and standards - in AI policy and regulations of the significant areas that in the AI policy has to be focused on is setting clearly of ethics and frameworks Wong, 2021). AI systems are being integrated into it is to ensure they in that and The establishment of ethical guidelines for AI is being in to issues of and transparency of AI decision-making from ethical is a need for safety standards in AI development 2018; de Almeida et al., 2021). AI has the risks to it through its and of therefore, that comprehensive safety be put into is a among policymakers of the need for setting standards that ensure AI systems will be and risks with Another critical affecting policy the economic implications of AI is a to and economic risks associated with the impact of on economic are that a of policies that can be put in to in and for new in the are being by both governments and also the to have policies to support while ensuring that AI economic are This for AI research and for ethical AI and regulation practices in the technology industry. governance of the regulation of AI AI systems are on of In the AI policy will further around data and to and data Another relevant in AI policy is increasing on and et al., The development of AI is not to it is a to different regulations across problems for and are guidelines and standards for AI. are across to help and AI technologies In the will be more and regarding the governance of AI that will to a much more approach by AI. policy will also be by the role of AI in critical such as healthcare, finance, and transportation these have specific regulatory requirements so that the development and applications of safety, and in example, in healthcare, the of AI for and stringent and to to algorithms and systems need regulations of and and policymakers are out sector-specific guidelines for with particular issues related to AI in key of policy AI public and widespread AI technologies require in their fairness, and policymakers are to focus on to public trust through the nature of in AI decision-making processes and for and It AI ethics public and broad in AI to trust and ensure that such technologies the for public Another of AI policy in the will be and the of AI, public understanding of its implications is AI policies at in at at public at making of the and risks associated with these emerging technologies. This the of AI into the and for learning to each to in an and approaches in AI policy and its further regulation will have to be of innovation for more regulations to new and that regulatory sandboxes and that AI technologies within controlled These approaches provide regulators with the to and policies to ensure that regulations relevant and a technological The industry regulation and policy of AI has at the focus of governments and organizations As AI is in from healthcare, finance, and - the of regulatory has been more a The AI Act is the European to a risk of AI systems, and a of for high-risk applications. This framework to ensure transparency, accountability, and for in the for the regulation of AI. increased regulatory activity in the United States of the National AI Act to in AI by and China is developing AI to ensure technological with stringent data and algorithmic transparency regulation. By a they to secure innovation as well as AI However, The around how can AI be in such a way that it not in AI both as an economic and a a is developing on approaches to regulations related to AI across in to out an for AI ethics and by agile approaches to regulation, will be as and to these to a on AI and its for security, and Artificial Intelligence and its on in Capraro, & (2024). The impact of artificial intelligence on inequalities and policy artificial legal and and challenges. of the Society and de & S. Artificial intelligence a framework for and AI in the European Union. of the current legal framework of the 2021). of and J., & Goldsmith, artificial for a In of the on AI, and Society (pp. M., & Artificial intelligence risks and algorithmic regulation. European of Regulation, S. of Artificial International Hoffmann-Riem, Artificial intelligence as a challenge for and regulation. artificial intelligence, Lauterbach, Artificial intelligence and Policy, and J., & (2024). for and ethical by under the European Artificial Intelligence & & Kaplan, Artificial to and & M., & in artificial it is and how it AI & Paramesha, M., Rane, N. L., & Rane, Artificial Machine Deep and in and A Paramesha, M., Rane, & Rane, Artificial AI at Paramesha, M., Rane, & Rane, Artificial intelligence in challenges, and ethical at Paramesha, M., Rane, & Rane, artificial intelligence such as in transportation A comprehensive at Paramesha, M., Rane, & Rane, data artificial intelligence, machine learning, of and for intelligence. at N. and regulation of artificial intelligence and framework and at Rane, N. key in of and similar artificial intelligence in manufacturing, finance, retail, and industry. Rane, & Rane, deep learning with machine technological and challenges. at Rane, & Rane, Artificial Intelligence of and for and Rane, & Rane, Artificial intelligence and machine learning for and sustainable and at Rane, & Rane, Artificial intelligence, machine learning, and deep learning for analysis in to and at Rane, & Rane, Artificial Intelligence and Machine in and a and a Rane, Paramesha, M., & Rane, Intelligence through Artificial A at Rane, & Rane, Artificial Intelligence and Machine in and A and International Rane, Paramesha, M., & Rane, Artificial Machine and Deep for A International Rane, Paramesha, M., & Rane, Machine and Deep for A of and International should artificial of the Society and J., J., M., & The approach to artificial an analysis of policy, ethics, and regulation. and policies in artificial intelligence, & data issues in artificial intelligence deployment. Taeihagh, of artificial intelligence. Policy and the of privacy, and in the artificial intelligence age. L., & and ethical of artificial on European policy Policy, N. and in artificial intelligence of Policy & Rademacher, artificial intelligence Wong, and regulation of artificial intelligence. In Artificial Intelligence for International at (pp. International Publishing. the A framework for artificial intelligence. &