We’ve reached a point in the evolution of AI where the questions we ask need to shift toward the uncomfortable reality of what we are doing wrong, rather than focusing solely on how we are getting things right. The merits of AI, and the speed with which the technology is creating growth opportunities, are unquestionable. However, it has become far more important, especially from a research perspective, to examine the impact of AI on us — not only from a productivity standpoint but also a cultural one.
To that end, Workday has released a new research paper titled “The Copy/Paste Economy: Why Task-Oriented AI is Failing the Enterprise.” This report explores the disconnect between AI ambition and AI reality in the workplace, particularly because workers spend too much time copying and pasting insights from one system to another.
AI Can’t Be a Bolt-On
To gather data for the report, Workday surveyed 6,100 active AI users from companies with 500+ employees across finance, HR, IT, and operations. One of the most significant findings is that while the goal of AI integration into workplace systems was aimed to free up employees to focus on valuable and engaging work, 82% of those surveyed said that, in reality, most of their time is spent as translators of AI outputs, by which they need to copy and paste information between systems.
Here’s a quote from a director of a construction company highlighting that statistic: “My day often feels busy but not genuinely productive when I’m pulled into constant coordination tasks and system‑related issues that interrupt focused, high‑value work.”
While Workday’s research rebuts the narrative that employees are scared of AI, 97% of respondents said they like their role, far too much time is spent on work that AI systems could undertake on their behalf.
Seventy-seven percent said that their work required them to “reconcile conflicting data from different tools,” while 70% were tasked with re-entering the same information into different tools or systems. In terms of productivity impact, 20% of those surveyed said they lost over seven hours a week to this manual integration, with 25% of IT professionals saying this effort defined their workday.
Embedded AI
Despite the obvious bottlenecks identified by Workday’s research, the report also showed the potential of AI technologies in the workplace when embedded properly. 83% of users surveyed said that AI is “already improving how their day-to-day work feels,” while 54% reported accelerated productivity.
While users are clearly confident in the work they are doing and in AI’s capabilities to support and improve it, when AI is used in isolation, the role of human operators is to manage this disconnect. The real magic happens when AI is fully integrated into critical, underlying systems and workflows — doing the work they don’t need to do.
The key is embedding AI into core processes, eliminating the need for employees to manually transfer information from one system to another. Here’s a quote from the report that summarises this perfectly:
“Workday’s research confirms this: nearly all employees say AI increases their confidence in decisions when they trust the underlying system and data. This is especially true of operations and IT professionals: 89% say AI becomes more valuable when it runs on a trusted operational core.”
Ultimately, Workday lays out a three-stage battle plan:
- Fix the Foundation: Modernize core systems and data to enable AI to produce trusted and consistent outputs.
- Integrate AI: Incorporate AI into end-to-end processes rather than isolating tasks. This approach gives humans the agency to take ownership of each step in the process.
- Design for Embedded AI: Create intelligent and invisible AI that is delivered seamlessly within familiar tools and systems as part of the workflow.
AI Agent & Copilot Summit is an AI-first event to define opportunities, impact, and outcomes with Microsoft Copilot and agents. Building on its 2025 success, the 2026 event takes place March 17-19 in San Diego. Get more details.
