As technology evolves from assisting people to executing tasks and making decisions within defined parameters, business leaders have stressed the urgent need for organisations to adopt agentic artificial intelligence (AI) with a clear purpose, strong governance frameworks, and sustained human oversight.
The call was made during a recent virtual webinar organised by the EY Nigeria Alumni Association as part of EY Global Alumni Week.
The event, themed ‘Leading with Confidence in the Age of Agentic AI,’ examined the implications of autonomous AI systems for operating models, decision-making, customer service, and workforce transformation.
The webinar featured EY technology experts, including Wilfred Mamah, partner and digital transformation services leader at EY West Africa, and Tomiwa Adefokun, senior manager, technology consulting at EY. The session was moderated by Bisi Onasanya, former group managing director of FirstBank.
According to Mamah, agentic AI represents a significant shift from software that merely supports users to systems capable of pursuing goals, planning tasks, using tools, and executing approved actions.
‘The defining shift is not simply better content generation; it is the delegation of tasks, decisions, and actions to software,’ he said. ‘Organisations should therefore treat agentic AI as a transfer of defined decision rights rather than as an enhanced chatbot.’
Mamah highlighted immediate applications across banking, insurance, and public services, including credit and service orchestration, claims triage and settlement support, and citizen-service fulfilment.
He noted that such systems can collect evidence, coordinate approvals, flag exceptions, and communicate next steps efficiently.
He further observed that the broader impact of agentic AI may be organisational rather than purely technological, as work shifts from departmental silos toward business outcomes, routine execution gives way to exception management, and decisions become more continuous and responsive.
For his part, Adefokun advised organisations to calibrate an agent’s level of autonomy based on factors such as risk, reversibility, confidence, and business impact. Depending on the circumstances, an AI agent may recommend or prepare an action, execute it with approval, act independently while notifying a human, or escalate uncertain and high-impact matters for human intervention.
He urged business leaders to clearly define the outcomes assigned to AI agents, the decisions they can make independently, their escalation points, the individuals accountable for their actions, and the data and systems they are permitted to access.
‘The future is not autonomous AI operating without people,’ Adefokun said. ‘It is responsible human-agent collaboration, where technology provides speed and scale while people provide purpose, judgement, and accountability.’
Also speaking at the webinar, Ashish Bakhshi, partner and head of markets at EY West Africa, challenged participants to translate the discussions into practical action within their organisations. He noted that successful adoption would depend not on deploying the highest number of AI agents, but on responsibly redesigning valuable work and earning stakeholder trust.
The speakers collectively maintained that responsible deployment requires a named business owner, clearly defined authority and data-access limits, human oversight of sensitive actions, traceable activities, continuous monitoring, and formal procedures for approving, updating, suspending, or retiring AI agents.
They also warned against the risks associated with excessive permissions, manipulated instructions, unauthorised actions, leakage of confidential information, compromised third-party integrations, weak audit trails, and uncontrolled proliferation of AI agents.