AI Training Tops Planned Enterprise Investments for 2025

What happened

A new survey from digital experience platform provider Liferay finds that employee training in artificial intelligence is the most commonly planned investment among organisations over the coming year. The research, reported by No Jitter, points to a shift from experimentation with agentic AI toward more structured rollouts and trials, with companies now prioritising workforce readiness alongside the technology itself.

The findings suggest that as agentic AI — systems capable of taking autonomous action rather than simply generating content — moves from pilot projects into wider use, organisations are recognising that tooling alone will not deliver value without parallel investment in how employees understand and work alongside these systems.

Why it matters

The survey signals a maturing phase in enterprise AI adoption. Early rollouts of generative and agentic AI have often focused on procurement and platform selection; the emphasis on training suggests leaders are now grappling with the harder, slower work of embedding these tools into daily operations and decision-making.

For digital transformation leaders, this is a useful corrective. Agentic AI changes workflows, role boundaries and accountability in ways that pure automation did not — employees need to know when to trust an agent's output, when to intervene, and how to escalate exceptions. Treating training as a planned investment, rather than an afterthought, reflects a more realistic view of what adoption actually requires.

The Renascence take

The instinct to fund AI training is right, but training is frequently treated as a one-off event rather than a behavioural change programme — and that is where most rollouts quietly stall.

Most organisations will budget for "AI training" as a knowledge-transfer exercise — a course, a certification, a rollout guide — when the real barrier is behavioural, not informational. Employees don't under-use agentic AI because they lack instructions; they under-use it because trust, accountability and habit are unresolved. A customer-obsessed operator should design training around the moments employees will actually hesitate — when to override an agent, who owns an error, what "good enough" looks like — rather than around features. Skip that, and the investment shows up in the training budget but not in the customer experience.