Exploring the Potential and Limitations of Artificial Intelligence and Machine Learning in Improving Decision-Making Processes in Various Industries

Exploring the Potential and Limitations of Artificial Intelligence and Machine Learning in Improving Decision-Making Processes in Various Industries

Adeleye Dupe Ayesha., Prof. Abd. Rahman Ahlan., Dr. Najhan Muhamad Ibrahim., Bello Bariz Holaitan., Yusuf Abdulkadir., Alijafar Umar., Abdulrahman Usman Liman

International Islamic University, Malaysia

DOI: https://doi.org/10.51584/IJRIAS.2025.100700129

Received: 17 July 2025; Accepted: 23 July 2025; Published: 21 August 2025

ABSTRACT

The study assesses the application of AI and ML in decision-making across four sectors: healthcare, banking, manufacturing, and retail. To measure the utilization in actual situations, the form of a mixed method unites expert interviews and questionnaires with simulated case studies. AI and ML serve to improve the accuracy of predictions, automate routine tasks, and customize processes. But there are issues with providing assurance information is unbiased and with model transparency and ethics concerns. Towards this end, the study advocates for implementing industry-specific regulations and human-machine collaboration in realizing optimal benefits optimally in a fair manner while limiting associated risks. Thus, the analysis’ top priority is still in adopting responsible and ethical approaches under each sector’s framework.

For this research, the outcome emphasizes the ways in which AI and ML continuously increase predictive precision, enable predictive maintenance, enable drudgery to be made rational in routine, and drive customer experience customization. That is, there remain some challenges: data bias, model building ambiguity, actual or perceived ethical issues confusing AI, or assumed over-reliance on computer software designed to shatter some form of human oversight. The present study presents new avenues of operational transparency, data integrity, and human-machine interface to facilitate co-working for best utilization of fullest potential of such technologies. It also posits that these strategies must be custom and agile to each sector based on its technology maturity level, regulatory conditions, or situations. In a way, this study validates profit-generating applications of AI-ML decision-making and yet expresses its concern for ethical, responsible, and inclusive ways of doing so that ultimately push some of the risks to human well-being and accumulate the behemoth potential of humankind.

INTRODUCTION

Despite all that, AI and ML are subject to the quality of data, transparency, and ethics.  This study seeks to assess their uses, the positives, and the negatives objectively in a variety of disciplines. While the application of AI and machine learning is for the efficiency of organizations and product personalization in factories and stores, and risk analysis and appraisal in areas such as health and economics, the arrival of the former transformed the way business decisions are made.  The systems rapidly evaluate enormous amounts of data, revealing patterns that guide decisions related to strategy and implementation (Mishra, 2022). Exploring how such innovations might be used in a variety of sectors to optimize operations and attain previously unheard-of levels of precision and effectiveness has become increasingly popular as a result. The capacity of AI and ML to handle and comprehend massive quantities of statistical information at an astounding rate holds the promise of enhancing ways to make decisions. Such systems can find hidden connections, trends, and structures that can escape human study by utilizing sophisticated algorithms. Organizations are able to make data-driven choices, improve operational effectiveness, and react quickly to changing marketplace conditions thanks to this information.

Different businesses have started to accept AI and ML as crucial tools for decision-making. For example, in healthcare, these advancements can aid in illness diagnosis, outcome for patients’ prediction, and treatment plan optimization. Financial organizations use AI and ML systems to evaluate risk, spot fraud, and offer specialized investment guidance (Carbone, 2022). AI- powered solutions in the industrial sector improve production schedules, anticipate maintenance needs, and improve supply chain processes. The applications include a wide range of industries where AI and ML have shown they can transform choice-making, including retail, travel, energy, and several more.

However, it’s critical to understand how AI and ML are constrained when used in methods of decision-making These technologies are very dependent on the type and volume of data they receive. Inaccurate or biased datasets might provide erroneous results and confirm preexisting biases. Additionally, it is still difficult to comprehend and explain AI and ML models, particularly in situations when significant decisions must be made and candor and responsibility are essential. Another factor that needs careful attention is finding a balance between mechanization and judgment from people (Carbone, 2022). The foundation for a fuller understanding of the revolutionary potential of these developments is laid by this investigation of the promise and constraints of AI and ML in enhancing how decisions are made in many sectors. Organizations can unlock the full promise of AI and ML while reducing risks by thoroughly examining their use cases and tackling the difficulties they provide. Although the process of utilizing such technologies to maximize decision-making is still in progress, its results promise to change businesses and equip experts with insightful knowledge for a better and more productive future.

LITERATURE REVIEW

Artificial Intelligence (AI) and Machine Learning (ML) have recently shown tremendous promise in a wide range of industries. Automation of work and increased productivity are two of AI’s and ML’s main benefits. These advances may greatly increase performance and lower human error by simplifying company operations and improving the distribution of resources. Additionally, data analysis with AI and ML is possible at a scale that is beyond human capacity. They may glean insightful information from huge databases, enabling enhanced forecasting and more educated choices (Tiwari, 2023). Another significant advantage is the possibility for modification and customization, since AI and ML approaches can provide customizable settings for learning, tailor healthcare medication, and tune consumer experiences.

Despite their enormous promise, AI and ML also have a number of drawbacks and difficulties. Statistical-related problems, such as worries about data dependability, impartiality, and quality, offer a substantial barrier. It is crucial to solve these issues for fair and impartial AI and ML technologies since biased data might produce biased outcomes (Lee, 2023). There are also cultural and ethical repercussions, such as prospective job loss, security hazards with regard to data, and privacy issues. All of these considerations must be carefully taken into account in order to ensure the ethical growth and implementation of AI and ML systems.

Scientists and professionals must also contend with the comprehension and simplicity of AI and ML models. Because some models function as “black boxes,” it might be challenging to comprehend how they make judgments. Concerns concerning trust, responsibility, and the capacity to spot and address biases or errors are brought up by this lack of understanding. The broad use of AI and ML is also hindered by technical limitations, such as having to use flexible and successful algorithms and processing power needs (Tirkolaee et al., 2021). Professionals and academics are continually developing ways and future initiatives to lessen these restrictions. To assure data quality and remove bias, this involves enhancing data collection and pretreatment methods. Effective AI and ML development requires the creation of ethical principles. Further, efforts are being made to improve the understanding of models and provide comprehensible artificial intelligence methodologies so that consumers can comprehend how AI systems make decisions (Gupta et al., 2022). In order to resolve constraints and make guarantees that the system properly supplements the abilities of humans, cooperation between artificial neural networks and human specialists is also essential.

AI and ML’s Development in Decision-Making: The development of AI and ML has made it possible for these technologies to be integrated into decision-making procedures. Early applications were mostly rule-driven, but recent developments in deep learning as well as neural networks have made it possible to create more complicated models that can handle complex and unstructured data. AI and ML have been adopted by sectors including banking, medical care, industrial production, and marketing to help decision- making (Paper, Wu and Jin, 2022). Algorithms are excellent at examining enormous volumes of data, finding trends, and offering insights that traditional analysts would miss. With the use of this capacity, businesses can make decisions with confidence and in real time, improving operational effectiveness and future planning (Lee, 2023).

Forecasting Analytics: AI and ML can forecast upcoming patterns and outcomes by using past data and using sophisticated predictive models. With the use of these predictive capabilities, businesses can better manage their resources, their inventories, and their interactions with customers, eventually increasing their profitability (Gupta et al., 2022).

Automation and Effectiveness: Artificial intelligence (AI) and machine learning (ML) technologies have the ability to automate complicated and repetitive processes, freeing up employee resources for higher-value and innovative projects (Mishra, 2022). Industries may improve production, eliminate mistakes, and simplify procedures, all of which contribute to the success of the business as a whole.

Personalization and Consumer Experience: Through suggestions and targeted commercials, AI and ML offer individualized customer experiences in industries like retail and business (Walton, 2018). Businesses may adjust their services and increase satisfaction with and commitment to clients by researching client tastes and habits.

Restrictions and Difficulties: The accuracy of AI and ML models significantly depends on the quality and variety of input data. 3.1. Data Quality and Bias. Inaccurate or biased data can produce biased predictions and judgments, which perpetuates current imbalances and limits the potential of the technologies (Aung, Wong and Ting, 2021).

Transparency: Sophisticated AI and ML models sometimes lack accountability, making it difficult for participants to comprehend the reasoning behind certain decisions. This absence of openness may cause ethical problems and make it more difficult to comply with regulations (Chen et al., 2023).

Interaction between people and computers is essential for the effective implementation of AI and ML into decisions that are made. To get the best results, it is imperative to make sure that human knowledge supports machine-generated knowledge.

Safety and privacy: As AI and ML are used more often, industries are exposed to security risks and privacy violations. Maintaining confidence in such innovations requires protecting confidential data and making sure that data protection laws are followed.

Study Cases from Various Industries: 4.1. The making of choices in healthcare has the potential to be revolutionized by AI-powered medical diagnostics, modeling for epidemics of diseases, and drug-finding algorithms, enhancing patient results and allocating resources (Hilb, 2020).

Finance: powered by AI algorithms are used in the financial industry to supervise portfolios of securities, monitor market trends, and determine credit risk. These innovations improve the precision of choices and help make banking operations more effective (Ranson et al., 2023).

In the manufacturing sector, utilizes of AI and ML streamline logistics, anticipate equipment problems, and improve inspection and maintenance procedures (Carbone, 2022). These innovations result in less downtime, more efficient manufacturing, and lower costs.

Advertising and retail: Individualized suggestions, demand forecasts, and evaluation of sentiment enable specialized advertisements and control of stock. Customers are better engaged, and AI and ML increase sales.

AI and ML have enormous potential and promise, with considerable advantages in a variety of fields. However, it is crucial to recognize and deal with the constraints and difficulties brought on by these innovations (Tiwari, 2023). By doing this, professionals and scholars may work to develop and use AI and ML in a moral and ethical manner, maximizing their abilities while minimizing dangers and guaranteeing a good influence on society.

METHODOLOGY

This study uses the mixed-methods approaches design. The qualitative component included informal interviews with architects, experts, and technical practitioners from health care, finance, production, both of and retail. Thematic analysis was used to identify common themes. The statistical aspect included standardized surveys given to experts from the same industries. Inference and descriptive statistics were used to analyze survey responses and identify relationships. Simulated examinations were also utilized to assess AI/ML performance in taking decisions contexts. These included medical services diagnostics, expenditure forecasting, proactive maintenance in production, and retail demand forecasting. The results have been juxtaposed to established methods to assess accuracy, efficiency, and limitations.

In parallel, the qualitative stage entails distributing a well-organized questionnaire to a broad variety of experts to get their opinions on the prevalence, advantages, and drawbacks of AI and ML in making choices within their particular industry. Statistical analyses that are qualitative as well as inferential are performed on the survey results. In general, inferential statistics assist in identifying links between variables, providing for a more thorough comprehension of quantitative information, whereas descriptive figures are used to describe the sample and evaluate survey results.

Additionally, ethical issues are crucial at every stage of the study process. All participants are asked for their informed permission to ensure they are aware of the study’s objectives, methods, and any dangers. To protect the privacy of participants and uphold data security, precautions are taken. With the help of both the numerical rigor of survey data and subjective insights from qualitative interviews, this mixed-methods technique offers a thorough examination of the study issue. The combination of these methods provides a thorough knowledge of how AI and ML affect various sectors’ decision-making procedures.

(ML) in enhancing the way decisions are made across diverse sectors. The tests sought to statistically assess how AI and ML initiatives affected the results of their choices.

Study designs were created to mimic decision-making circumstances in many sectors using actual-life situations. For instance, tests were designed to examine how artificial intelligence (AI) diagnostic tools affected the precision and timeliness of medical treatments in the healthcare industry. Similar studies were carried out in the financial sector to evaluate how well ML algorithms predicted stock market changes and improved investment plans.

These projects produced predictions and suggestions using data gathered from genuine industrial scenarios using AI and ML algorithms. To ascertain if AI and ML offered greater insights, efficacy, or accuracy, the results were contrasted against conventional decision-making techniques.

The trials also sought to address the shortcomings and difficulties of AI and ML, such as selectivity in data and comprehensibility of model results. The tests added real-world proof to the continuing discussion on the practical consequences of integrating AI and ML in making decisions across a variety of businesses by methodically altering parameters and evaluating various scenarios.

CONCLUSIONS AND DISCUSSION

The investigation of artificial intelligence’s (AI) and machine learning’s (ML) capabilities and constraints in strengthening procedures for making decisions across numerous sectors has revealed a landscape full of both potential and challenges. The results of this study provide insight on the transformational potential and real-world applications of merging AI and ML by revealing the intricate connections between technology breakthroughs, workplace dynamics, and ethical issues.

Implementing AI and ML technology has shown to significantly improve ways of making decisions. Artificial intelligence, predictive analytics, automation, improved personalization, and information-driven decisions stand out as important advantages across sectors. For instance, AI- powered diagnostic systems in the medical field have demonstrated astounding precision, enabling the early diagnosis of diseases and the development of exact treatment plans. Predictive repair in the industrial sector has reduced downtime and enhanced the utilization of resources, increasing production efficiency. These accomplishments highlight how AI and ML have the power to completely alter current paradigms for making decisions.

Challenges And Limitations

But along with these encouraging developments, difficulties and restrictions need to be carefully taken into account. The significance of understanding and accessibility is highlighted by ethical problems resulting from prejudiced algorithms and opaque decision-making processes. Furthermore, for the best decision results, the mutually beneficial partnership between human knowledge and machine-generated knowledge must continue. The study emphasizes the crucial part that knowledgeable human assistance plays in improving AI and ML capabilities and fixing their flaws.

Cross-Industry Variations

The report also emphasizes how crucial it is to understand how AI and ML uptake and effect vary across industries. The adoption of AI in the medical industry has been sparked by customized medicine and illness prediction, as opposed to the finance industry, which uses AI for risk identification and prediction. These variances represent distinct particular to an industry possibilities and difficulties, demanding sector-specific AI and ML initiatives.

The Human-Machine Nexus

Recognizing the complex relationship between AI, ML, and individuals making decisions is essential. An environment of cooperation where AI supplements human insight without replacing it is essential for effective adoption. The human-machine nexus is crucial, particularly in situations when moral judgment, original thought, and empathy are required. Achieving this balance guarantees that ML and AI are useful tools in the course of making choices.

FUTURE DIRECTIONS AND ETHICAL CONSIDERATIONS

Future study and creation need to tackle the ethical issues inherent in such technologies as companies continue their journey toward implementing them. It is crucial to maintain equity, responsibility, and openness in algorithms that make choices. The possible disruption of certain employment rolls calls for proactive initiatives to reskill employees and leverage AI and ML’s advantages to advance socioeconomic development.

Exploring the promise and constraints of AI and ML to improve taking decisions across many sectors is a multifaceted path. This study highlights the transformational potential and advantages while also highlighting the complex ethical, interpretive, and creative aspects. Businesses may use the potential of AI and ML to enable improved knowledge, effective, and significant decisions to be made by integrating the technologies strategically and responsibly.

REFERENCE

1. Aung, Y.Y.M., Wong, D.C.S. and Ting, D.S.W. (2021) ‘The promise of artificial intelligence : a review of the opportunities and challenges of artificial intelligence in healthcare’, 0(August), pp. 4–15.
2. Carbone, M.R. (2022) ‘When not to use machine learning : A perspective on potential and limitations’, MRS Bulletin, 47(9), pp. 968–974. Available at: https://doi.org/10.1557/s43577- 022-00417-z.
3. Chen, Y. et al. (2023) ‘Human-Centered Design to Address Biases in Artificial Intelligence Corresponding Author ’:, 25(Figure 1), pp. 1–10. Available at: https://doi.org/10.2196/43251.
4. Gupta, A. et al. (2022) ‘Artificial intelligence : A new tool in surgeon ’ s hand’, (Ml). Available at: https://doi.org/10.4103/jehp.jehp.
5. Hilb, M. (2020) ‘Toward artificial governance ? The role of artificial intelligence in shaping the future of corporate governance’, Journal of Management and Governance, 24(4), pp. 851–870. Available at: https://doi.org/10.1007/s10997-020-09519-9.
6. Lee, D.K. (2023) ‘Application of Artificial Intelligence in Oral Cavity Cancer : A Review Artificial Intelligence ( AI ) Applications in Oral Cavity Cancer’, 34(2), pp. 23–26.
7. Mishra, S. (2022) ‘Artificial Intelligence : A Review of Progress and Prospects in Medicine and Healthcare’, 4(1), pp. 1–23.
8. Paper, P., Wu, T. and Jin, R. (2022) ‘The Limitations of Deep Learning in Achieving Real Artificial’, pp. 1–5.
9. Ranson, J.M. et al. (2023) ‘Harnessing the potential of machine learning and artificial intelligence for dementia research’, Brain Informatics [Preprint]. Available at: https://doi.org/10.1186/s40708-022-00183-3.
10. Tirkolaee, E.B. et al. (2021) ‘Application of Machine Learning in Supply Chain Management : A Comprehensive Overview of the Main Areas’, 2021(Ml).
11. Tiwari, R. (2023) ‘The Integration of AI and Machine Learning in Education and its Potential to Personalize and Improve Student Learning Experiences’, (2020), pp. 1–11. Available at: https://doi.org/10.55041/IJSREM17645.
12. Walton, P. (2018) ‘Artificial Intelligence and the Limitations of Information’. Available at: https://doi.org/10.3390/info9120332.