Governance, Risk and Trust Determine Scale

As organisations move from pilots to deployment, governance concerns are becoming increasingly important. Data security and lack of explainability emerge as the most significant risk concerns associated with agentic AI.

Interestingly, workforce displacement is less emphasised than operational and governance risks, suggesting that executives are increasingly focused on trust, accountability and decision quality. These concerns are amplified by today’s still-limited governance maturity.

Biggest Risks of Moving Away from an FTE-Heavy Model

Less than half of GBS organisations implemented robust governance policies around the use of agentic AI to date. Many organisations still rely on informal practices (33%) or have yet to establish appropriate governance structures altogether (22%).

This governance gap presents a significant risk. Agentic AI systems operate with greater autonomy than traditional automation technologies, increasing the importance of clear accountability, monitoring frameworks and oversight mechanisms.

Without these foundations, organisations may struggle to scale beyond isolated pilots regardless of technological capability.