Workday used its Workday Rising conference in Las Vegas this week to roll out a trio of AI products aimed squarely at finance departments: new "autopilot agents" meant to automate routine accounting, payroll, and hiring workflows, an expanded Trusted Agent Platform for building custom AI agents, and a new partnership with Anthropic to integrate agentic tools more deeply into Workday's software. The announcements come under co-founder Aneel Bhusri, who reclaimed the CEO role in February after previously stepping back from day-to-day leadership.
Workday CFO Zane Rowe used the moment to push back on how most companies are judging whether their AI spending is paying off. In comments at the conference, Rowe argued that simply measuring how much more output a team produces misses the point, because higher volume doesn't automatically translate into better business outcomes. "It's not just about finding efficiencies," Rowe said. "It's about people actually doing the work that they thought they were going to do when they signed up to join a company."
A different scorecard for AI spending
Rather than a single return-on-investment number, Rowe said finance chiefs should be tracking a broader set of signals: whether AI tools are leading to stronger decisions and higher-quality work, whether they're expanding what an organization can actually take on, whether they're translating into better outcomes for customers, and whether employees report feeling more engaged rather than reduced to supervising software. That framing puts Rowe in a small but growing group of finance executives arguing that traditional productivity metrics undersell, or outright mislead on, what AI tools are actually doing to a company's output.
Testing the tools on Workday's own books
Workday is using its own finance organization as a proving ground for the approach. The company's finance transformation team is testing AI products internally before they reach customers, and its tax team has been building out roughly 10 different AI agents for tax-related workflows, with Rowe acknowledging that not all of them are expected to make it into production. Workday is also expanding its Decision Intelligence tool, which models workforce scenarios and forecasts how different staffing or spending decisions would ripple through the business, giving finance teams a way to test choices before committing to them. Taken together, the announcements position Workday's finance arm not just as a seller of AI tools to other companies' CFOs, but as a live test case for the harder question Rowe is raising: what actually counts as success when a company starts handing routine work to AI agents.
