Technology Solutions & Practices Shared Interest Group

Author:

Monomita Nandy, PHD, FHEA, FAcSS

Brunel University of London

Monomita.nandy@... (corresponding author)

Suman Lodh, PhD, FHEA

Kingston University London

S.Lodh@...

Abhishek Singh Rajpurohit

Founder Vitaminai & Voice of crypto

abhishek@...

As the accounting profession navigates the digital frontier, Agentic AI is redefining the boundaries of strategic decision-making. Unlike conventional AI, Agentic AI systems demonstrate autonomy, contextual awareness, and goal-oriented behavior- making them powerful allies in management accounting. Moreover, the Agentic AI monitors its environment and can adjust its actions with unexpected changes without continuous human prompts. So the algorithm breaks the actions into smaller steps that helps in structured approach of management accountants as any sudden change may impact budgeting, cost control, and performance measurement.

For academic researchers, this opens new pathways in behavioral finance, ethical modelling, and predictive analytics. Agentic AI can simulate managerial dilemmas, test policy interventions, and co-create adaptive financial models. For instance, an AI agent embedded in a budgeting platform might autonomously flag anomalies, suggest cost-saving alternatives, and refine forecasts based on historical and real-time data.

Policy makers can leverage Agentic AI to model fiscal outcomes and regulatory impacts. Imagine a system that simulates the long-term effects of tax reforms or public spending initiatives enhancing transparency and responsiveness in governance.

From a business perspective, adopting Agentic AI in management accounting drives significant competitive advantages by optimizing resource allocation and enhancing profitability. Enterprises can achieve up to 20-30% reductions in operational costs through proactive anomaly detection and automated scenario planning, allowing finance teams to focus on high-value strategic initiatives rather than routine tasks

In industry, management accountants are already seeing the benefits. A recent use case involved a global firm deploying Agentic AI to monitor working capital across subsidiaries. The system not only identified inefficiencies but proposed tailored interventions-such as renegotiating supplier terms or adjusting inventory based on local conditions and strategic priorities.

However, integration is not without challenges. Data integrity and governance, algorithmic bias, maintaining traceability and auditability and professional accountability must be addressed. The profession must also rethink roles, skills, and ethical frameworks to ensure AI complements-not replaces human judgment. This will ensure the trust in agentic AI.

Call to Action:

We invite academic institutions, professional bodies, and industry leaders to collaborate in shaping the future of Agentic AI in management accounting. Let us co-develop training pathways, pilot innovative models, and build systems that uphold transparency, sustainability, and strategic value. If you would like to collaborate with us in organising training for your clients, conducting research, developing policy documents, or seeking strategic assistance for your business, please feel free to get in touch.

The future of accounting is not just digital- it's agentic. Let's lead it together.

#AgenticAI #ManagementAccounting #IMA #Sustainability #AccountingInnovation #DigitalTransformation #StrategicFinance #AIethics #AcademicIndustryCollaboration

------------------------------
Monomita Nandy
Professor in Accounting and Finance
Brunel University of London
UK
------------------------------

Hi Monomita Nandy

Thanks for articulating the differences between traditional AI and agentic AI.

AI's decision-making process-especially in the context of Agentic AI-can be broken down into a structured, goal-driven framework that mimics human strategic thinking but operates autonomously and adaptively. Briefly mentioning the process below:

---

Agentic AI Decision-Making Process

1. Perception (Input Gathering)

• What happens: The AI continuously monitors its environment-internal systems (e.g., ERP, financial databases) and external factors (e.g., market trends, regulatory updates).
• Tools used: Sensors, APIs, data connectors, real-time feeds.
• Example: Detects a sudden spike in raw material costs or delayed supplier payments.

2. Contextual Understanding

• What happens: The AI interprets the data in context, considering historical patterns, business rules, and strategic goals.
• Tools used: Machine learning models, semantic analysis, business logic engines.
• Example: Recognizes that the cost spike is seasonal and correlates with supplier performance history.

3. Goal Formulation

• What happens: Based on its understanding, the AI sets or aligns with predefined goals (e.g., reduce costs, optimize cash flow, maintain service levels).
• Tools used: Rule-based systems, reinforcement learning, optimization algorithms.
• Example: Sets a goal to reduce procurement costs by 10% without affecting delivery timelines.

4. Planning and Simulation

• What happens: The AI breaks down the goal into actionable steps and simulates various scenarios to evaluate outcomes.
• Tools used: Scenario modeling, digital twins, Monte Carlo simulations.
• Example: Simulates renegotiating supplier contracts vs. switching vendors vs. adjusting order volumes.

5. Decision Execution

• What happens: The AI selects the optimal action and initiates it autonomously or recommends it to human decision-makers.
• Tools used: Workflow automation, robotic process automation (RPA), decision engines.
• Example: Automatically sends a proposal to renegotiate terms with a supplier and updates the procurement plan.

6. Feedback and Learning

• What happens: The AI monitors the results of its actions and learns from outcomes to refine future decisions.
• Tools used: Feedback loops, supervised/unsupervised learning, performance analytics.
• Example: Learns that renegotiation led to cost savings and improved delivery times, reinforcing similar future decisio

Each of these steps are independent processes by themselves, they are either triggered waiting for an input from the previous process or acting in parallel while feeding the next process. This kind of orchestration is a key part of the agentic AI, which either spawns multiple agents based on need and collects the feedback from those agents to generate a cohesive reply to a situation.

------------------------------
Eashwar Seshadri CMA
Director/Manager
IBM India Ltd
Norristown PA
United States
------------------------------

Thank you so much Eashwar! 

------------------------------
Monomita Nandy
Professor in Accounting and Finance
Brunel University of London
UK
------------------------------

Original Message:
Sent: 10-26-2025 01:53 AM
From: Eashwar Seshadri
Subject: Agentic AI and the Strategic Evolution of Management Accounting

Hi Monomita Nandy

Thanks for articulating the differences between traditional AI and agentic AI.

AI's decision-making process-especially in the context of Agentic AI-can be broken down into a structured, goal-driven framework that mimics human strategic thinking but operates autonomously and adaptively. Briefly mentioning the process below:

---

Agentic AI Decision-Making Process

1. Perception (Input Gathering)

• What happens: The AI continuously monitors its environment-internal systems (e.g., ERP, financial databases) and external factors (e.g., market trends, regulatory updates).
• Tools used: Sensors, APIs, data connectors, real-time feeds.
• Example: Detects a sudden spike in raw material costs or delayed supplier payments.

2. Contextual Understanding

• What happens: The AI interprets the data in context, considering historical patterns, business rules, and strategic goals.
• Tools used: Machine learning models, semantic analysis, business logic engines.
• Example: Recognizes that the cost spike is seasonal and correlates with supplier performance history.

3. Goal Formulation

• What happens: Based on its understanding, the AI sets or aligns with predefined goals (e.g., reduce costs, optimize cash flow, maintain service levels).
• Tools used: Rule-based systems, reinforcement learning, optimization algorithms.
• Example: Sets a goal to reduce procurement costs by 10% without affecting delivery timelines.

4. Planning and Simulation

• What happens: The AI breaks down the goal into actionable steps and simulates various scenarios to evaluate outcomes.
• Tools used: Scenario modeling, digital twins, Monte Carlo simulations.
• Example: Simulates renegotiating supplier contracts vs. switching vendors vs. adjusting order volumes.

5. Decision Execution

• What happens: The AI selects the optimal action and initiates it autonomously or recommends it to human decision-makers.
• Tools used: Workflow automation, robotic process automation (RPA), decision engines.
• Example: Automatically sends a proposal to renegotiate terms with a supplier and updates the procurement plan.

6. Feedback and Learning

• What happens: The AI monitors the results of its actions and learns from outcomes to refine future decisions.
• Tools used: Feedback loops, supervised/unsupervised learning, performance analytics.
• Example: Learns that renegotiation led to cost savings and improved delivery times, reinforcing similar future decisio

Each of these steps are independent processes by themselves, they are either triggered waiting for an input from the previous process or acting in parallel while feeding the next process. This kind of orchestration is a key part of the agentic AI, which either spawns multiple agents based on need and collects the feedback from those agents to generate a cohesive reply to a situation.

------------------------------
Eashwar Seshadri CMA
Director/Manager
IBM India Ltd
Norristown PA
United States
------------------------------

Monomita, thank you for sharing your insightful perspective on Agentic AI and the strategic evolution of management accountants. I would like to respond to your discussion on this topic.

"Policy makers can leverage Agentic AI to model fiscal outcomes and regulatory impacts. Imagine a system that simulates the long-term effects of tax reforms or public spending initiatives enhancing transparency and responsiveness in governance."

Policymakers today, face the challenge of making decisions in an environment defined by complexity, uncertainty, and competing priorities. Agentic AI offers a transformative tool to navigate this landscape by enabling the simulation of fiscal outcomes and regulatory impacts with unprecedented depth and precision. Unlike traditional economic models, which often rely on static assumptions, Agentic AI systems can dynamically adapt new data, test multiple scenarios, and reveal long-term consequences of policy choices in real time.

Imagine a government considering a major tax reform. With Agentic AI, policymakers could simulate not only the immediate revenue implications but also downstream effects on household consumption, business investment, and regional economic disparities over decades. Similarly, when evaluating public spending initiatives such as infrastructure investment or healthcare expansion, Agentic AI could model ripple effects on employment, productivity, and social equity. This capacity to forecast across interconnected domains enhance both transparency and accountability, as citizens and stakeholders can see how decisions are likely to play out under different conditions.

Beyond forecasting, Agentic AI can foster responsiveness in governance. By continuously integrating new economic indicators, demographic shifts, and behavioral data, these systems allow policies to be stress-tested and recalibrated as circumstances evolve. This creates a feedback loop where governance becomes more adaptive, evidenced-based, and resilient.

Ultimately, leveraging Agentic AI in fiscal and regulatory planning is not about replacing human judgment but augmenting it. By equipping policymakers with richer insights and clearer foresight, Agentic AI can help bridge the gap between short-term political pressures and long-term societal well-being-laying the foundation for more informed, transparent, and responsive governance.

------------------------------
Karen Gayle FMAA, EA
Accountant
Bronx NY
United States
------------------------------

Agree with you Karen! We should encourage policy makers to start using recent technology 

------------------------------
Monomita Nandy
Professor in Accounting and Finance
Brunel University of London
UK
------------------------------

Original Message:
Sent: 10-27-2025 02:03 PM
From: Karen Gayle
Subject: Agentic AI and the Strategic Evolution of Management Accounting

Monomita, thank you for sharing your insightful perspective on Agentic AI and the strategic evolution of management accountants. I would like to respond to your discussion on this topic.

"Policy makers can leverage Agentic AI to model fiscal outcomes and regulatory impacts. Imagine a system that simulates the long-term effects of tax reforms or public spending initiatives enhancing transparency and responsiveness in governance."

Policymakers today, face the challenge of making decisions in an environment defined by complexity, uncertainty, and competing priorities. Agentic AI offers a transformative tool to navigate this landscape by enabling the simulation of fiscal outcomes and regulatory impacts with unprecedented depth and precision. Unlike traditional economic models, which often rely on static assumptions, Agentic AI systems can dynamically adapt new data, test multiple scenarios, and reveal long-term consequences of policy choices in real time.

Imagine a government considering a major tax reform. With Agentic AI, policymakers could simulate not only the immediate revenue implications but also downstream effects on household consumption, business investment, and regional economic disparities over decades. Similarly, when evaluating public spending initiatives such as infrastructure investment or healthcare expansion, Agentic AI could model ripple effects on employment, productivity, and social equity. This capacity to forecast across interconnected domains enhance both transparency and accountability, as citizens and stakeholders can see how decisions are likely to play out under different conditions.

Beyond forecasting, Agentic AI can foster responsiveness in governance. By continuously integrating new economic indicators, demographic shifts, and behavioral data, these systems allow policies to be stress-tested and recalibrated as circumstances evolve. This creates a feedback loop where governance becomes more adaptive, evidenced-based, and resilient.

Ultimately, leveraging Agentic AI in fiscal and regulatory planning is not about replacing human judgment but augmenting it. By equipping policymakers with richer insights and clearer foresight, Agentic AI can help bridge the gap between short-term political pressures and long-term societal well-being-laying the foundation for more informed, transparent, and responsive governance.

------------------------------
Karen Gayle FMAA, EA
Accountant
Bronx NY
United States
------------------------------