IIBA membership provides you with exclusive access to Member Articles.
Agentic AI and the Business Analyst: How to Shape, Govern, and Deliver AI Agents That Work
Key Takeaways
- Agentic AI redefines automation: Unlike traditional tools, agentic AI systems act autonomously, requiring business analysts to define clear goals, boundaries, and escalation rules
- Governance is key: Effective agentic AI requires proactive governance, including guardrails, audit trails, and operational controls, starting from the design phase
- Autonomy levels matter: Defining and managing autonomy levels ensures agents operate safely and align with organizational risk tolerance
- Business analysts as "agent architects": Business analysts play a central role in shaping, governing, and delivering AI agents that are practical, trustworthy, and scalable
- Operational readiness drives success: Success depends on testable requirements, monitoring plans, and user training to ensure agents deliver real outcomes without compromising accountability
Business analysts are the architects of agentic AI, shaping trusted, scalable solutions through clear goals, strict controls, and proactive governance.
Disclaimer: The views and opinions expressed in this article are those of the author and may not reflect the perspectives of IIBA.
In many organizations, the first wave of generative AI looked like a faster search bar (even though it’s nothing like a search at all). You asked a question, you received a response, and you decided what to do next.
Agentic AI changes that pattern. Instead of only generating text, an agent can plan steps, call tools, interact with systems, and complete tasks on a user's behalf. That shift is powerful, but it’s also risky. A system that can act needs clearer boundaries than a system that only advises.
The scale of this shift is significant.
Gartner projects that by the end of 2026, approximately 40 per cent of enterprise applications will embed task-specific AI agents, up from less than five per cent at the start of 2025. The agent market itself is projected to grow from roughly $8 billion to over $50 billion by 2030.
Deloitte's
2026 Tech Trends study found that while 30 per cent of organizations are exploring agentic options and 38 per cent are piloting solutions, only 11 per cent are actively using them in production.
The ambition is there, but the execution gap is wide. And in that gap sits the business analysis professional.
When autonomy becomes part of a workflow, requirements expand beyond screens and fields. We now need to specify goals, constraints, decision rights, safety checks, escalation rules, audit trails, and what success looks like in operations. Agentic AI is a problem of clarity, accountability, and design of work. Those are business analysis problems.
This article offers a practical playbook: what to define, which artefacts to produce, how to write testable requirements, how to manage risk and governance, and how to prepare teams for go-live.
