ERP Governance

MCP is emerging as the protocol layer that lets AI agents discover and call enterprise tools, data sources, and workflows. For ERP leaders, the opportunity is lower integration friction. The risk is that access control, auditability, and cost discipline move closer to the agent boundary.

Enterprise leaders, while ambitious in their generative AI strategies, face a significant gap between their AI aspirations and actual production adoption, as illustrated by Mukesh Kumar's insights on targeted automation in ERP systems addressing specific, high-impact problems.

The EU AI Act delay eases some near-term compliance pressure, but ERP teams still need to build governance evidence around how AI is used across business processes.

ERP cost overruns stem more from post-purchase management and governance issues, like underutilized licenses and ongoing subscription growth, rather than the initial license pricing itself.

ERP programs falter not from inadequate change management but from early weak structural decisions, with the need for 'Transformation Capability Management' (TCM) to ensure organizations build durable abilities rather than depend on external partners, as traditional change management alone cannot address underlying capability, capacity, and governance challenges.

Sovereign AI is gaining traction in Asia as governments, like India and Singapore, prioritize local control over AI infrastructure and data governance, prompting enterprises to adapt their systems and strategies to comply with varying regulations and infrastructure conditions across the region.

Microsoft’s 2026 Data Security Index shows generative AI adoption accelerating faster than data security controls, exposing governance and visibility gaps across enterprise environments.

Organizations are modernizing ERP superficially, with cloud adoption outpacing essential governance and strategy, leading to inconsistent deployment, limited returns, and an urgent need for structured integration and support for achieving measurable value.