Page 89 - CMA Journal (Mar-Apr 2025)

T T T

3) Tax Compliance: From data gathering to filing made. Black-box AI models that can’t explain their
returns, the AI ensures timely and compliant tax logic face skepticism—especially in regulated
processes, while staying updated with rule changes. industries.
4) Continuous Auditing: Instead of periodic reviews, 3) Cybersecurity and Privacy: Agentic AI handles
Agentic AI enables real-time monitoring of sensitive data—from payroll to audit trails. Ensuring
transactions and internal controls, flagging issues as data protection, encryption, and access control is
they arise. non-negotiable.
5) Audit Evidence Collection: The AI pulls required 4) Evolving Regulations: Jurisdictions worldwide are
documentation from emails, systems, and logs still shaping laws on AI use in financial
without manual request lists or follow-ups. decision-making. Compliance with both financial
and data governance regulations must be
6) Fraud and Risk Analysis: By analyzing patterns
maintained.
across large datasets, Agentic AI can uncover
potential fraud indicators or policy violations that 5) Skills Gap: Finance professionals need to
might escape human notice. understand how to interpret AI outputs, validate
algorithms, and manage exceptions. Upskilling is
7) Client Advisory Services: With routine work
crucial to bridge this gap.
handled by AI, accountants can offer deeper
insights and tailored recommendations, using 6) Cost and Change Management: Initial
AI-generated data summaries and forecasts. implementation can be resource-intensive.
Organizations need to demonstrate ROI and
8) Due Diligence Support: In mergers and
manage cultural resistance to change.
acquisitions, AI agents quickly assess financial and
operational risks by scanning large volumes of A New Role for Accountants and Auditors
documents and metrics.
Rather than being replaced, professionals will be
Who’s Leading the Adoption? redefined. The future accountant is not a bookkeeper
but a strategist—someone who validates AI outputs,
While widespread implementation is still maturing, the
interprets complex data relationships, ensures ethical
momentum is clear:
compliance, and translates insights into business value.
• Big Four firms are developing proprietary agentic
Auditors will become continuous assurance
systems for continuous auditing, dynamic risk
professionals, leveraging AI for deeper, real-time
assessments, and smart tax engines.
coverage and shifting their focus to judgment,
• Startups and fintechs are creating modular AI stakeholder engagement, and regulatory liaison.
agents that integrate with accounting platforms like
Agentic AI marks a revolutionary point in the evolution
Xero, QuickBooks, and SAP.
of accounting and auditing. It brings speed, scale, and
• Corporations are exploring agentic tools in internal intelligence to tasks that were once repetitive and
finance functions—particularly in shared service manual. While the road to full adoption requires careful
centers and global reporting teams. planning, ethical considerations, and technical
readiness, the destination promises immense value.
In most cases, these are hybrid
implementations—where humans oversee and validate Those who embrace agentic AI early—integrating it
AI outputs. But as trust grows and performance thoughtfully into processes and teams—will lead the
improves, autonomy is gradually increasing. future of the profession. In this future, man and machine
work in harmony: AI drives the tasks, and humans drive
Challenges and Considerations
the strategy.
Despite its potential, the path to agentic AI is not About the Author: Alia Noor is an Associate Partner at Ahmad
without hurdles: Alagbari Chartered Accountants and also holds the position of
Director VAT in the LEA Consultant & Director Program at Alif
1) Data Quality and Integration: AI is only as good as Technologies. She has been associated with EY and Unilever
the data it receives. Many organizations still struggle Pakistan for external and internal audits. She is the founder and
with fragmented systems, inconsistent coding owner of “xpertsleague” a free learning platform. She holds
practices, and incomplete datasets. professional qualifications, FCMA, CIMA, and MBA along with
Certifications in Oxford Fintech Program, Digital Mindset, COSO
2) Explainability and Trust: Professionals and Enterprise Risk Management, COSO Internal control, and GCC

regulators need to understand how decisions are VAT Diploma.

ICMA’s Chartered Management Accountant, Mar-Apr 2025 87