Will the AI black box arrive in time?

by underscience

Regardless of ethics or apocalyptic predictions, artificial intelligence (AI) is already a central component of economic and institutional decision-making. The International Journal of Intelligent Systems Design and Computing study goes beyond industry-specific analysis of cutting-edge AI to provide a detailed framework for how different AI tools are being adopted.

The main takeaway from the analysis is that while AI technology is widely used across sectors, organizations still do not have a strategy that allows them to integrate AI in a way that balances innovation and accountability.

AI includes so-called machine learning to recognize patterns in data, natural language processing that can interpret human language, and generative tools to produce text, images, video, computer code, and other output. All these tools are changing many fields, from medical diagnostics to processing industrial and financial data to producing hit pop songs and their accompanying videos.

Education and business operations are undergoing similar changes. Adaptive learning platforms in education tailor course materials to the way individual students learn. In retail and logistics, AI is being used to improve supply chains, manage inventory, and personalize the customer experience. In the legal world, law enforcement agencies are using AI to assess crime scenes and weigh evidence, and judges are using these tools to summarize concluding remarks from vast briefs.

One of the most pressing issues highlighted by this study is data privacy, as AI systems rely on large amounts of sensitive and personal information. Then there is the concept of algorithmic transparency, where we are losing the ability to understand how a particular AI system reaches a particular decision. In fact, many of the most advanced AI models currently function as essentially black boxes. That means its internal processes simply can’t be interpreted…you’ll probably have to rely on another AI to do so. This lack of transparency can undermine trust in high-stakes situations such as medical diagnoses and judicial decisions.

To address this issue, researchers have proposed a framework based on stakeholder theory. This framework emphasizes the importance of all parties affected by decisions that AI may make. They stress that in a business context, organizations should not only focus on efficiency and profits, but also consider the interests of employees, customers, regulators, and society as a whole when implementing AI. Of course, this can only be achieved through governance, regulation and ethical obligations.

Idemudia, EC (2025) “Effect and impact of artificial intelligence on multiple fields and sectors”, Int. J. Intelligent System Design and Computing, Vol. 3, No.3/4, pp. 254–274.
DOI: 10.1504/IJISDC.2025.152183.