Volume 4 (2025-6)

This paper is an excerpt from ‘Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work, and Life’ by Bornet et al. (DOI: 10.1142/14380). It examines how agentic artificial intelligence (AI) has emerged from the convergence of two previously independent technological streams: the rapid evolution of large language models (LLMs) and the maturation of automation technologies, from early robotic process automation (RPA) to contemporary intelligent automation. The excerpt traces the historical development of both streams, illustrating how advances in neural networks, transformerbased language models and workflow automation have progressively removed barriers between ‘understanding’ and ‘doing’. It discusses the significance of early LLM-based agent frameworks, such as Modular Reasoning, Knowledge, and Language (MRKL), ReAct and Toolformer, which introduced mechanisms for tool use, reasoningaction loops and external system interaction. The excerpt also provides an overview of the fastgrowing agentic AI market, outlining emerging categories of platforms and agent types. As a whole, the excerpt provides foundational context for understanding how agentic AI systems have become technically feasible and why they represent a new phase in AI capability. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: AI; artificial intelligence; GenAI; agentic AI; corporate digital responsibility; services marketing; service management

The swift emergence of agentic artificial intelligence (agentic AI) signifies not merely a new trend in automation but a fundamental transformation in the design, governance and execution of work. Unlike preceding waves of machine learning and generative AI (GenAI), which predominantly offered recommendations or generated content, agentic AI introduces software agents capable of perceiving context, making decisions, executing tasks across various systems and learning over time. Procurement serves as an exemplary domain for demonstrating this paradigm shift. It operates at the confluence of finance, supply chain, legal, risk, environmental, social and governance (ESG) and business stakeholders, managing complex, rules-intensive, high-value decisions. This paper posits that agentic AI will profoundly overhaul the procurement operating model, transitioning it from function-centric workflows to an AI-native hub that orchestrates autonomous agents, human experts and interconnected data. It outlines the practical aspects of agentic AI, explains why procurement is central to its adoption, and describes how capabilities such as autonomous sourcing, negotiation, onboarding and compliance monitoring reshape processes, roles and decision-making rights. Furthermore, it proposes an operating model blueprint for AI-native procurement, encompassing governance, risk management and policy systems. It examines the skills, behaviours and organisational structures necessary for effective human–AI collaboration. Ultimately, it offers a pragmatic roadmap for chief information officers (CIOs) and chief procurement officers (CPOs) and considers implications for other enterprise functions. This framework enables the design of procurement organisations that leverage the value of agentic AI while maintaining control, resilience and trust.
Keywords: agentic AI; enterprise procurement; operating model transformation; AI-native organisations; autonomous agents; AI governance; human–AI collaboration

Agentic artificial intelligence (agentic AI) systems capable of autonomous decision making are increasingly being deployed in global trade compliance to enhance speed, accuracy and responsiveness. Applications such as automated customs classification and real-time sanctions screening offer operational efficiencies, but also introduce significant legal, operational and ethical risks — particularly under the European Union Artificial Intelligence Act (EU AI Act), which becomes fully applicable in August 2026. This paper examines how agentic AI reshapes the trade compliance landscape and why its deployment increases exposure to costly errors and regulatory liability. It argues that the opacity, scalability and decision-making autonomy of agentic systems amplify existing compliance challenges and complicate explainability, auditability and error remediation in the absence of robust governance frameworks. The paper further explores how compliance functions must evolve from output-focused monitoring toward supporting AI life cycle governance, human oversight and continuous post-market monitoring, while emphasising that primary responsibility for AI governance and accountability should rest with practitioners and corporate leadership rather than being delegated to trade compliance teams. Drawing on a comparative analysis of regulatory approaches across the EU, US, China and selected other jurisdictions, the paper advances a human-centric oversight model designed to preserve accountability, traceability and legal certainty. It concludes that integrating governance, risk management and operational controls is essential to enabling responsible use of agentic AI in international trade compliance. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: agentic AI; autonomous systems; trade compliance; EU AI Act; human oversight; AI governance; customs classification; supply chain compliance; post-market monitoring; risk management

Agentic artificial intelligence (agentic AI) represents a fundamental shift from reactive automation to autonomous decision-making systems. These systems can perceive environments, reason for complex scenarios, and take independent action to achieve specified objectives. Unlike previous generations of generative AI (GenAI) that respond to user prompts, agentic AI systems operate with inherent autonomy, managing end-to-end workflows without constant human intervention. The retail industry stands at the forefront of agentic AI adoption. Organisations are deploying autonomous agents in customer service, inventory management, dynamic pricing and personalised commerce; however, this autonomy introduces unprecedented risks. AI agents can hallucinate false information, be manipulated through adversarial attacks, make decisions misaligned with organisational values and propagate errors across interconnected systems. This paper examines critical dimensions essential for responsible agentic AI deployment in retail: strategic business applications that create differentiated value, and governance frameworks that establish trust through transparency and explainability. Organisations that effectively integrate these dimensions will harness agentic AI’s transformative potential while fostering stakeholder confidence and ensuring regulatory compliance. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: agentic AI; autonomous agents; retail technology; supply chain management; inventory optimisation; transparency; AI governance; responsible AI deployment; retail transformation

The integration of agentic artificial intelligence (agentic AI) into the software development life cycle (SDLC) and business process transformation presents a critical tension between speed and rigour. Organisations, in their rush to automate, risk overlooking the disciplined analysis essential for strategic alignment, value realisation and risk mitigation. This paper argues that, far from diminishing, the role of business analysis (BA) and business analysts (BAs) becomes more vital in the age of agentic AI. The BA must evolve beyond traditional requirements gathering to become a ‘context architect’ and ‘strategic enabler’, bridging business intent, AI capability, and human-in-the-loop (HITL) adoption. Effective analysis is necessary to prevent ‘AI slop’ and ‘AI drift’, ensure compliance and design sustainable hybrid human–AI systems. The paper proposes the needed shift in thinking about the BA’s role and expanding BA competencies, concluding that elevating business analysis to a strategic capability is essential for achieving resilient, responsible and value-driven intelligent enterprise transformation. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: agentic AI; AI; business analysis; business analyst; AI governance; human-in-the-loop (HITL); requirements analysis

This paper examines security incidents affecting artificial intelligence (AI) coding assistants and enterprise AI agents during 2025, providing security teams with practical guidance for risk assessment and mitigation. The paper synthesises notable vulnerabilities and public advisories affecting major platforms including GitHub Copilot, Cursor, Amazon Q Developer, Microsoft 365 Copilot and Claude Code, with a focus on enterprise-relevant impact rather than exhaustive coverage. Three vulnerability patterns recur across platforms: prompt injection enabling privilege escalation, inadequate authentication at trust boundaries, and insufficient isolation between AI operations and sensitive resources. These patterns enabled remote code execution, credential theft and data exfiltration. Organisations deploying AI agents require updated security controls including agent inventory management, configuration hardening and vendor security assessment. The recurring patterns across diverse tools indicate the need for new forms of trust infrastructure and governance beyond product-by-product patching. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: AI security; prompt injection; agentic AI; enterprise risk; AI governance

Agent artificial intelligence (AI) automates AI using the reasoning power of large language models (LLMs) to make decisions without human intervention. A governance framework is needed to create transparency of decision lineage in multi-agent systems to offer fixes throughout the life cycle of agent development and deployment to ensure agency and safety for humans. This paper outlines five approaches to agent governance and compares the different approaches to offer guidance on this important topic across multiple industries. This includes tracking for metrics,1 security risk from agents calling application programming interfaces (APIs),2 data governance,3 human-in-the-loop approaches,4 and rules to test adversarial attacks for edge cases.5 This paper is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: agentic AI; AI agents; AI; agentic AI governance; mixture of agents; multi-agent systems; responsible AI

The rapid evolution of artificial intelligence (AI), particularly in the form of agentic systems, has introduced unprecedented capabilities for autonomous decision making in various sectors, including financial services.1 These agentic AI systems, characterised by their ability to act independently and learn from interactions, promise to revolutionise credit decisioning by enhancing efficiency, accuracy, and risk assessment. The autonomous nature of agentic AI, however, necessitates a robust framework for managing inherent risks, including issues of accountability, bias, and compliance, especially within the highly regulated financial industry. This paper proposes the development and implementation of agentic AI Risk-Aware Credit Decisioning Sandbox (RA-CDS) as a critical mechanism to foster responsible innovation while mitigating potential systemic risks. This controlled environment would allow for the rigorous evaluation of AI agents’ performance and adherence to ethical guidelines within credit decisioning processes.2 This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: agentic AI systems; credit decisioning; financial risk management; AI governance; regulatory sandbox; explainable AI; autonomous decision making

Enabling non-technical domain experts to use automation by employing simple low-code/no-code interfaces has always been considered key to driving adoption of technology for improving business value. In the emerging era of agentic artificial intelligence (AI), where intelligent systems act autonomously to achieve defined goals, this empowerment takes on renewed significance. Organisations worldwide are now seeking not only to equip employees with AI-driven tools but also to enable them to build their own intelligent agents that enhance productivity and innovation. This evolution marks a new phase of human–machine collaboration: humans contribute deep, often tacit domain knowledge and intuition, while AI systems provide analytical depth and adaptive decision making. As low-code/no-code tools democratise application creation, non-technical business users are emerging as key co-architects of autonomous, goal-driven AI agents. Success in this space, however, is not automatic. Businesses need to identify the right use cases, choose flexible and secure platforms, and put in place strong governance models to manage risk and ensure trust. This paper introduces a reference capability model to implement a citizen development approach that offers a structured framework for organisations to evaluate their current maturity, identify the capabilities required for scaling and reinforce areas of existing strengths. The resulting assessment serves as a strategic roadmap, guiding organisations towards successful and sustainable deployment of agentic AI. Like every transformative technology before it, citizen development powered by AI is inevitable; it is not a question of if, but when. Early, deliberate and well-governed adoption will position organisations to achieve sustainable success in this rapidly evolving technological era. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: AI governance; citizen developers; low code no code; AI agent framework

The concept of software agents has been discussed for many decades with the vision that agents ‘inhabit some complex dynamic environment, sense and act autonomously in this environment, and by doing so realize a set of goals or tasks for which they are designed’, as defined by Pattie Maes in 1995. An illustrative example is the case of car navigation systems, which estimate a best route according to the preferences of the driver, adapt to traffic news with optimised routing and inform the user which way to go. Only recently, the tremendous development of large language models (LLMs) triggered a new wave of excitement about so-called ‘AI agents’, which are based on a LLM core enhanced with interfaces to tools and orchestration of software elements, which are programmed ex-ante. These AI agents combine two limitations: LLMs are statistical estimators for most probable ‘next best tokens’, and tools such as interfaces or orchestration have to be programmed traditionally, ie knowing which problem has to be solved. There are, however, narratives that they can transform businesses operations and increase productivity by making decisions without humans, optimising processes and adapting instantaneously to new situations. Following a brief review of the actual technical implementations, this paper scrutinises this issue from three perspectives: the so-called t-bench (‘tool–agent–user’, ie benchmark with retail, airline and telecom customer support systems), first tests with more sophisticated AI agents (such as Anthropic’s project ‘Vend’ in 2025), and real-world tests of agent-like LLM applications for financial services (such as extraction of environmental, social and governance [ESG] parameters from corporate reports). The paper concludes that the usability of AI agents depends on a quality–resource analysis for every individual use case: there is a difference between a macro-economic analysis of a set of ESG reports versus an individual decision for corporate ESG-linked lending. While tailored benchmarks for ‘fine-tuned’ LLMs to solve dedicated problems such as maths text questions are quite impressive, the current experiences with real-world cases do not support the vision that AI agents would rewrite the rules of business, but indicate a possible new Solow paradox. Originally, Robert Solow wrote in 1987: ‘You can see the computer age everywhere but in the productivity statistics.’ This paper understands this paradox as analysed by Daron Acemoğlu et al. in 2014: that IT-using industries show no additional productivity gains, in contrast to the view that IT is making workers redundant and ‘automates’ performance increase, despite the continuously increasing output of the IT-producing industry. Today, this paradox might reappear for AI-agent-using versus AI-agent-producing sectors of economy. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: AI agents; generative AI; large language models; statistical estimators; Solow paradox

In recent years, the combination of increasingly accessible artificial intelligence (AI) and the rapid pace of technological development has led to numerous analyses of its potential to make significant changes to labour market dynamics, industrial organisation and macroeconomics. Nevertheless, a considerable gap remains in our knowledge regarding the actual impact of AI. This paper, which focuses on a unique survey conducted by Israel’s Central Bureau of Statistics, aims to examine this critical issue by focusing on the Israeli case as part of the global discussions around these issues. The paper addresses three key themes: the effect of AI adoption in businesses on the workforce composition, the influence of AI on disparities between economic sectors, and the barriers to broader and deeper implementation of AI technologies in the private sector. Regarding labour substitution, it appears that although businesses in Israel already make relatively widespread use of AI to perform tasks previously carried out by workers, the effect on overall employment levels remains relatively limited. It is yet to be seen whether a more significant substitution effect will occur in the longer term, or whether the size of the effect was overestimated in previous studies. When it comes to the differential diffusion rates between industries, Israel stands out in international comparison. Israel’s high-tech sector demonstrates much higher levels of AI implementation compared to other industries. Relatedly, unequal adoption rates between the geographical centre of the country and its periphery are also significant. The Israeli case thus offers further evidence that AI adoption in the labour market may deepen existing disparities, especially in countries where innovation-based industries attract high proportions of highly skilled employees. Finally, among the different implementation barriers in traditional industries, the lack of knowledge and understanding of AI emerges as a major impediment to adoption. This phenomenon may again point to the human capital imbalances in dual economies. The paper concludes by discussing potential policy takeaways. Such a policy might include promoting the integration of AI in traditional industries, offering more robust lifelong learning frameworks, as well as relevant education reforms. Considering the rapid technological developments in AI, it is also important to continue monitoring this issue regularly with agile yet consistent empirical instruments. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: artificial intelligence; dual economy; comparative survey analysis; labour market; Israeli economy

Data science offers significant potential for leveraging traditional and emerging data sources, improving statistical processes and enabling complex data analysis in central banking. It also facilitates secure sharing of granular datasets while protecting sensitive information. There are various challenges, however, including the need for robust IT infrastructure, organisational barriers and limited quality of secondary sources, which constrain their use for official purposes. Fortunately, central banks are well equipped to address these issues, not least because of their expertise as both data producers and users. Looking ahead, enhancing data management, promoting interoperability, investing in modern data platforms and fostering structured exchanges of experiences across stakeholders can be essential to unlock the full potential of data in today’s modern societies and effectively support central banks’ public mandates. This paper examines how central banks can harness data science to improve data management, overcome organisational and technical challenges, and implement strategies that unlock the full potential of data for policy and operational purposes. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: data science; data access; data sharing; central banks; official statistics; innovation

This paper explores the strategic integration of artificial intelligence, blockchain technology and dynamic capabilities (DC) in the pharmaceutical sector. The research utilises a mixed-methods approach that integrates quantitative benchmarking and case studies of six organisations to elucidate the operational realities of technology adoption. The primary findings suggest that a substantial financial investment does not necessarily result in a positive return on investment. Instead, the analysis identifies distinct strategic archetypes, ranging from ‘foundation builders’, characterised by the presence of substantial yet underutilised infrastructure, to ‘lean innovators’, who demonstrate high efficiency with minimal expenditure. A central conclusion of this paper is that DC is the critical enabler that allows companies to translate technological potential into tangible value. Pharmaceutical organisations with robust DCs exhibit considerably higher project success rates and more adeptly leverage their skilled workforce to enhance operational efficiency. The paper provides a data-driven framework for benchmarking these strategic approaches, demonstrating that competitive advantage hinges on aligning technological investments with robust organisational capabilities. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: artificial intelligence; innovation; healthcare; pharmaceutical; life sciences; operations management; dynamic capabilities; blockchain technology

Autonomous vehicles (AVs) cover the gamut from robotaxis for consumers to autonomous mobility vehicles (AMVs) in factories, mines and warehouses, as well as everyday deliveries. They are all powered by artificial intelligence (AI), utilising computer vision and machine learning technologies to act as Edge AI mobility devices. This paper provides an update on the AVs of today, explains their improved technology stack and presents the challenges in building and deploying them in an ethical and value-generating way to help customers. The paper goes beyond presenting the technology to include practical use cases from the automotive, transport, hospitality and warehousing industries on how AMVs are utilising AI to automate workflow in an industry ecosystem of people, process and machines to create value from mobility. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: autonomous vehicles; AI; electronic control units; ECU; autonomous mobility robots; AMRs; safety

The current widespread use of artificial intelligence (AI) in law enforcement represents a paradigm shift towards intelligence-led, predictive and data-driven criminal management, resulting in unprecedented advances in operational efficiency and strategic decision making globally. While these developments characterise AI policing in some world nations, their use in resource-constrained, post-conflict situations such as the Sierra Leone Police (SLP) is unclear. Using the medium-sized criminal investigation department (CID) of the SLP, the study on which this paper is based examined the effect of AI on the operational performance of the SLP, considering the mediating and moderating effect of green management (GM). A survey research design was employed, following a quantitative and deductive approach. A total of 181 out of 196 staff members at the CID headquarters were sampled. Primary data was collected using a structured and validated questionnaire and analysed using inferential statistics, with Smart PLS-SEM as the analytical tool. The findings revealed a statistically significant positive effect of AI on the operational performance (OP) (β = 0.356, t = 4.054, p < 0.001) of the SLP CID while the moderating interacting effect (AI*GM) on OP is negative and statistically significant (β = –0.221, t = 2.517, p = 0.012). On the other hand, the indirect effect (mediating role) of AI on OP through GM is positive but insignificant (β = 0.067, t = 1.906, p = 0.057). The study concluded that deliberate investment in AI and conscious consideration of GM will be a strategic tool for boosting policing and small and medium-sized establishments. Therefore, this paper recommends deliberate investment, proper implementation and constant monitoring of the usage of AI tools alongside GM, which will boost the general OP of police establishments and specifically bring them up to speed in meeting the demand for global practices. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: artificial intelligenceamp;; criminal investigation department; green management; operational performance; Sierra Leone Police

This paper discusses the implementation and governance of artificial intelligence (AI) in central banks, drawing on a survey by the Irving Fisher Committee on Central Bank Statistics (IFC) of the Bank for International Settlements (BIS). AI has become strategically important for central banks, with the main question being how to use it effectively and responsibly to support their tasks and guide decision-making. Central banks’ experience shows that AI deployment is being complemented with adequate and robust governance, especially to ensure the development of effective user-focused applications and mitigate the associated risks. However, AI implementation also brings several IT challenges, including the pressing yet costly need to access more computational power. Looking ahead, making further progress on data management, governance and infrastructure, as well as ensuring that society can keep accessing high-quality reference information appear to be key priorities for making the most of AI opportunities.
Keywords: artificial intelligence; central banks; data governance; innovation; official statistics

Artificial intelligence (AI), especially Generative AI (GenAI), has been reshaping the business and public sector landscapes, forcing organisations to evaluate their digital strategies or risk falling behind their counterparts. However, technology alone does not guarantee success and return on investment, organisations must also establish core foundations. These fundamentals span capabilities such as organisational talent strategies, data management and governance and data ethics to harness the benefits of AI effectively. GenAI has been the catalyst for change. This paper explores the considerations for successful AI deployments and outlines the concepts that emphasise talent development, data governance, responsible innovation and well-aligned strategic objectives enabling leaders to position their organisations for sustainable AI-driven growth. This paper is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: artificial intelligence; generative AI; data governance; AI readiness; automation; data literacy; talent management; organisational culture; responsible AI; digital transformation

This study investigated the role of artificial intelligence (AI) in enhancing competitive advantage within Access Bank PLC in Lagos State, with strategic focus on Alimosho LGA. It examined the effects of AI technologies, including deep learning (DL), chatbots, robotic process automation (RPA), digital assistants and facial recognition, and assessed their impact on operational efficiency, customer success and strategic decision-making. The study employed a descriptive research design and surveyed 142 employees selected through simple random sampling. Data were collected through structured questionnaires and analysed using regression analysis to determine the significance of AI-driven innovations. The findings revealed that AI was found to significantly enhance operational efficiency (p < 0.01), improved customer personalisation (p < 0.01) and provided valuable strategic insights (p < 0.05), contributing to the bank’s competitive edge in market share and profitability. However, challenges such as high implementation costs, data privacy concerns and regulatory compliance issues were noted. The study recommends a strategic focus on responsible AI integration, employee upskilling and partnerships to mitigate challenges while maximising AI benefits. This research contributes to the understanding of AI’s transformative potential in banking and offers practical insights for stakeholders in the financial services sector. This paper is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: artificial intelligence; competitive advantage; operational efficiency; customer success; strategic decision-making; deep learning; chatbots; robotic process automation; digital assistants; facial recognition; banking

In cyber security, anticipating the worst is not paranoia — it is strategy. This paper introduces creative thinking frameworks designed to help leaders and security teams collaborate with generative AI (GenAI) to stress-test assumptions, identify opportunities, explore edge-case risks and uncover blind spots, often caused by cognitive biases, before they become breaches. Drawing inspiration from Stoic philosophy’s premeditation of evils — the practice of imagining future harms to build resilience — this paper repositions GenAI from its role as an operational efficiency tool to one where it serves as a creative partner in strategic planning. The paper explores how GenAI can enhance strategic imagination through various techniques that will expand security teams’ field of vision and be equipped with actionable tools for improving red teaming, scenario planning and creative problem-solving. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: generative AI; risk management; cyber security; red team; strategic planning

Artificial intelligence (AI) is revolutionising financial risk management by providing innovative tools to address diverse challenges. In two papers of my series ‘Machine Finance’, the author critically examines the integration of AI technologies into the financial sector, emphasising their applications in managing risks such as credit, market, operational and compliance. It highlights the transformative potential of AI-driven models and analytics, which offer enhanced precision, real-time monitoring and robust predictive capabilities. Additionally, this study explores the ethical and regulatory challenges associated with AI adoption, particularly within frameworks like the European Union (EU) Artificial Intelligence Act (AI Act). By comparing traditional risk management methods with AI-enabled approaches, this paper aims to provide practical insights into the future of financial stability and resilience through advanced AI-driven strategies. This paper is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
Keywords: artificial intelligence; AI; financial risk management; market risk; cyber risk; climate risk