Workday published research on July 13, finding that employees are using AI and feeling positive about its potential, but disconnected enterprise systems are still forcing workers to act as the manual glue between tools, teams, and business processes.
The research draws on a Harris Poll survey of 6,100 global professionals across HR, finance, IT, and operations who actively use AI at organizations with at least 500 employees. A report published in parallel with the results focuses on a 300-person regional subset from Indonesia, Malaysia, Singapore, and Thailand, which the report refers to collectively as ASEAN.
The findings challenge the idea that AI resistance is the main barrier to enterprise productivity. In the ASEAN sample, 97% of employees rated their day-to-day work positively, 86% said AI had improved their work experience, 69% said AI had reduced task completion time, and 67% said AI had accelerated work in a productive way.
The problem is what Workday calls the “copy/paste economy.” Employees are still spending large parts of the workday moving information between systems, reconciling reports, translating information between teams, re-entering data, and navigating administrative friction.
Workday found that 84% of ASEAN employees spend significant time moving information between systems, 84% coordinate or translate work between teams or business systems, 82% reconcile conflicting data or reports, and 75% re-enter the same information in multiple systems or tools.
The Embedded AI Gap
Workday’s research argues that AI delivers more value when it sits inside the systems and workflows that run the business, rather than operating as a standalone tool for isolated tasks. Yet, only 30% of ASEAN respondents said their organization has embedded AI directly into core workflows. Most organizations are still using AI around the edges for work such as drafting, summarizing, answering isolated questions, or accelerating individual tasks.
The time-savings gap is notable. In Asia Pacific and Japan, 65% of employees in organizations with AI embedded in core systems said AI reduced task time by 25% or more. Where AI is not used in core systems, that figure falls to 36%. Workday said that makes employees in AI-embedded organizations 1.8 times more likely to report meaningful time savings.
The report’s examples show where employees already want AI to operate:
- 55% use AI agents within core systems to monitor metrics and suggest actions
- 51% use them for onboarding
- 42% use them for HR and policy questions
- 41% use them for budgeting and forecasting
- 40% use them for financial close or reporting
- 38% use them for cross-departmental approvals.
Systems Friction Is the Adoption Problem
The research also complicates the trust narrative around enterprise AI. Only 12% of ASEAN employees cited lack of trust in AI as a barrier to adoption. More common blockers included uneven skills, training, or access to AI tools; AI outputs that misalign with rules, policies, or processes; rigid systems or workflows; poor data quality; too many approvals; unclear guidance; and unclear accountability.
That finding matters for ERP, HCM, and finance leaders because it shifts the AI adoption question away from employee sentiment and toward operating design. Employees may be willing to use AI, but they still need the surrounding systems, data, policies, and workflows to support reliable decisions.
Workday found that 93% of ASEAN employees said AI increases confidence in decisions when they trust the underlying system and data. The implication is clear: trust in AI depends heavily on trust in the enterprise systems and data the AI draws from.
The report also points to decision delays and data conflicts as persistent sources of friction. Three-quarters of ASEAN employees said missing or unclear information delayed decisions, and 72% said teams often disagree over whose numbers are right.
Get Our Free Weekly Newsletter
AI Needs Better Foundation
For enterprise leaders, the research suggests that adding more AI tools may not solve the productivity problem if core workflows remain fragmented.
Task-level AI can make individual work faster, but it may not remove the coordination burden created by disconnected applications, inconsistent data, unclear approval routes, or fragmented business processes. In some cases, bolt-on AI may simply help employees work faster inside the same broken operating model.
The more durable productivity opportunity is embedded AI that can work inside trusted systems, carry process context, understand approval chains, and support defined handoffs between humans and agents. Workday frames that shift as moving from AI that assists existing work to work redesigned around what AI and humans each do best.
Sponsor Industry‑Grade Research
What This Means for ERP Insiders
AI productivity depends on process integration, not tool volume. Enterprises can give employees more AI assistants, but the productivity ceiling remains low when people still have to copy data, reconcile reports, chase approvals, and translate information between systems. For CIOs, CHROs, CFOs, and operations leaders, the next priority is to evaluate whether AI is reducing workflow friction or simply speeding up isolated tasks.
Data consistency will decide whether embedded AI scales. AI agents need trusted business context, common definitions, clean process data, and reliable system access before they can support HR, finance, IT, and operations decisions. For ERP and HCM teams, the practical work starts with data governance, workflow design, and process ownership rather than another round of disconnected AI pilots.
Employee trust follows system trust. Workers are more likely to accept AI when it operates inside the systems they already rely on for payroll, financial close, onboarding, approvals, and reporting. For enterprise software vendors and transformation leaders, the market signal is that useful AI will need to become embedded, governed, and nearly invisible in the flow of work.
