Workday customers put AI costs, controls and workplace choices on the table

At Workday Rising, customers from Carlyle and Caterpillar discussed how they are weighing AI costs, setting guardrails and choosing where employees should encounter AI at work. Their practical point: measure AI’s added value separately from ordinary automation, and make approved use easier than everyone improvising with public chatbots.

Watch Desk analysis

What happened

Diginomica’s Jon Reed describes a customer panel at Workday Rising 2026, where attendees asked about return on investment, HR’s role in AI adoption and which tools should become employees’ “front door”. Carlyle’s Omer Tanvir said the value of a project needs to be quantified and executed transparently. He also urged companies to distinguish the gain from AI from the gain of automating a manual process.

Caterpillar’s Andy Konsky described work on hiring, self-service and payroll agents, alongside internal AI tools. He said the company was still considering its employee-facing approach, with Outlook and ServiceNow among the connectors being explored. Tanvir also described Carlyle standardising a legally approved version of AI use for its Talent 360 review process after employees had been entering feedback into ChatGPT with varying prompts. Workday was a Diginomica premier partner and paid for most of the author’s travel to the event.

Why it matters

These examples offer a useful test for businesses beyond Workday’s customer base: an AI project’s business case should not quietly claim credit for routine automation, and deployment needs rules people can follow. The panel’s comments are company representatives’ accounts, not independent measures of results, but they make the work involved more concrete than another promise of effortless productivity.

Our read

The best advice here is also the least glamorous: separate the spreadsheet columns for automation and AI, then make the approved workflow the usable one. “We have a policy” is a poor answer if staff are still left to guess. The panel offers experience, not a universal recipe, but at least it gives finance and IT something more useful to argue over than a token bill.

What to watch

• Whether Carlyle or Caterpillar publishes measurable results from the approaches described.
• How companies distinguish AI’s contribution from ordinary process automation.
• Whether employee-facing AI tools become simpler to standardise across different roles.

Discussion spark: Should companies require every AI project to prove value beyond ordinary automation before scaling it, or would that rule shut down experiments whose benefits take longer to show?

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.