Workday, a global workforce software company, is positioning enterprise AI apart from the wave of generic large language model (LLM) tools that have flooded the market in recent years. During a press conference in Dublin last week, the company made clear that “good enough” AI is not sufficient for human resources (HR), finance, and planning.
“Being inaccurate is not an option for us,” said Clare Hickie, Workday CTO for Europe, Middle East and Africa (EMEA), during a fireside chat titled “AI Visionaries”.
“It can’t be an option for us, and it never has been an option. [Being] 95% accurate in terms of how much you get paid, or the payroll to run, is not accurate. It’s actually wrong. To be 95% compliant – we can’t accept it. Our customers will never accept it. Everything that we do, everything that we deliver on, needs to be 100% accurate, and it needs to ensure there’s precision. It’s completely compliant, and, of course, it can be trusted.”
This position underpins Workday’s approach to pairing probabilistic AI, which operates on patterns and inference, with what the company describes as deterministic systems.
Kathy Pham, Workday VP of AI, said the company’s focus is on grounding AI in structured, enterprise data rather than relying solely on models trained on public information.
She said: “Most LLMs are trained on broad systems. For a while, two plus two in some of these systems was five, because if a system is trained on data where many sources say two plus two is five, then it will say it’s five.”
For Workday, that type of behaviour highlights a deeper mismatch between general-purpose AI and enterprise requirements. Systems trained on internet-scale data do not understand financial records, HR data, or planning processes, all of which require accuracy, consistency, and context.
“Where we see the value add for Workday is to bring in a deterministic perspective to what is probabilistic with AI,” she said, referring to the use of trusted business processes and governed data as the foundation for AI outputs.
The objective is not to replace probabilistic models, but to constrain them. By grounding AI in a unified data core and established workflows, Workday aims to ensure that outputs reflect verified enterprise data rather than inferred patterns.
Pham said enterprise systems must be designed with “the right permissions and controls and audit trails in place” so that users can’t access information that they are not supposed to.
“These agents need to understand real work, a deep understanding of how work happens, of these trusted business processes that we have in our systems.”
This requirement shapes how AI is embedded into day-to-day operations. Systems must understand how work is done, including decision paths, approvals, and security structures, rather than simply generating responses.
Workday AI strategy
Workday’s AI strategy is structured around four pillars, aimed at embedding AI directly into enterprise workflows rather than treating it as a standalone tool. These include:
- Workday as the front door to work – Workday is positioning its AI layer, Sana, as a central interface within the platform, where users can ask questions about payroll, HR processes, and regional complexities and receive responses within existing workflows.
- Transforming HR and finance with AI – The company is embedding AI into its core HR and finance products, which have been in use by customers for years, rather than introducing separate tools.
- Driving excellence with HR and finance – Workday is continuing to focus on HR and finance as core domains, applying AI within those areas.
- Unleashing the ecosystem with an open platform – The platform is being opened to customers, partners, and developers to build on top of Workday’s technology, governance, and security frameworks.
Pham said that even within a deterministic framework, enterprise AI must be deployed with different levels of control depending on the type of process being executed.
She told Gadget: “Something that’s lower risk, like scanning receipts for expense reports, versus using data to gather everything we’ve done the last six months to do performance reviews — there’s a higher risk there, because performance review data gets to then how employees are rated in the company, it gets to their compensation.”
That difference reflects how workflows are structured across the organisation, with financial reporting, payroll, and performance management requiring more tightly defined processes than routine administrative tasks.
“There’s nothing today that can look at an end-to-end workflow and really have a gut sense for how someone hires end to end, or how someone does end-of-quarter reviews of their financial records.”
For Workday, this highlights the challenge of applying AI across enterprise systems, where each workflow requires different levels of structure, control, and context.