AvePoint’s third annual State of AI Report highlights governance gaps and rising visibility issues as enterprises increase AI usage, signaling that trust and control are becoming critical to scale.

Even among the most confident organizations, 72% experienced an unauthorized data access incident in the past year, underscoring persistent security risks despite optimism.

The insights come from a survey of 750 enterprise leaders across the Americas, Europe, the Middle East, Africa, and Asia-Pacific, conducted with Osterman Research.

A core finding is that success with AI will hinge on building an operational trust layer—visibility, enforceable governance, and auditable outcomes—rather than relying on model capability alone.

Value from AI depends on readiness and a trusted data foundation, and organizations should adopt a comprehensive trust layer to enable scalable, auditable AI outcomes.

Investments highlighted include securing data used for AI training (top priority at 79.5%), adopting third-party governance tools to monitor agent actions, and expanding enforceable data foundations with auditing capabilities.

Organizations plan to invest in governance and control layers, including safeguarding training data and deploying AMP-like governance tools to ensure policy alignment.

In the next 12 months, top investments include securing AI training data (79.5%) and adopting agent governance tools to monitor agent actions for compliance.

The report stresses ongoing data security, privacy, and governance challenges as AI adoption accelerates, calling for actionable controls, auditing, and corrective mechanisms.

Nearly 90% of organizations delayed both agentic and generative AI deployments by an average of about six months due to data security and governance concerns.

Security governance delays are widespread, with 86% delaying generative AI deployments for about six months, highlighting the need for an enforceable trust layer of data visibility, governance, and auditability.

Altogether, deployment delays tied to data security and governance concerns emphasize the imperative for robust, auditable data governance before wide-scale AI rollout.