On May 28, 2026, Workday and Google Cloud announced an expanded partnership that puts Workday’s Sana Self-Service Agent inside Gemini Enterprise, makes Gemini the default model for Sana for Workday, and deepens the data connection between Workday Data Cloud and Google Cloud. It was easy to miss because it landed in a crowded late-May AI cycle. Two days later, it still looks worth covering because the move says something bigger about where enterprise agents are actually heading: away from standalone demos and toward the everyday work surfaces employees already live in.
This is not just a model-partnership story. Workday is trying to turn HR and finance support into a governed multi-agent workflow that can sit inside Gemini Enterprise, use Workday’s policies and approvals, and hand work across systems without asking employees to jump between portals, help desks, and back-office apps.
What Workday and Google Cloud actually changed on May 28
The May 28 announcement bundled several product and platform moves into one release. The practical takeaway is that Workday did not merely add another connector. It moved a live Workday agent into Google’s enterprise AI surface and tied that move to model choice, interoperability, and data architecture.
Key changes in the Workday-Google Cloud expansion
| Change | Why it matters |
|---|---|
| Sana Self-Service Agent is available in Gemini Enterprise | Employees and managers can handle Workday HR and finance tasks without leaving Gemini Enterprise. |
| Gemini becomes the default model for Sana for Workday | Workday is standardizing more of the reasoning layer for policy-heavy, multilingual, and nuanced internal requests. |
| A2A, A2UI, and MCP support | The partnership is aimed at real-time agent handoffs across systems, not a single isolated assistant. |
| Zero-copy link between Workday Data Cloud and Google Cloud Lakehouse | Teams can analyze HR and finance data and move from insight to action without copying sensitive data out of Workday. |
Workday said employees can use the Gemini Enterprise integration for tasks such as checking time-off balances, updating personal information, reviewing payslips, requesting leave, checking expense policies, and opening finance-related requests. Managers can also review team goals, approve timesheets in bulk, start performance reviews, and submit payroll input. For enterprises, that means the value proposition is not “ask an AI a question.” It is “complete a governed piece of work in the flow of work.”
Why this still matters after announcement week
The bigger pattern becomes clearer when this release is put next to Workday’s May 13, 2026 Microsoft 365 Copilot integration. In that earlier launch, Workday brought the same Sana Self-Service Agent into Microsoft 365 Copilot so employees could complete HR and finance tasks inside Microsoft’s workplace surface. The Google Cloud expansion on May 28 now makes the strategy harder to miss: Workday wants Sana to become a trusted execution layer that can sit inside the major enterprise AI surfaces people already use, while Workday stays the system of action, policy, and record.
That matters because many enterprise AI rollouts still stall at the same point. Employees know where they want help to appear, but core business logic, approvals, and sensitive data usually live somewhere else. Workday is trying to close that gap by keeping transactions and permissions anchored in Workday while exposing the agent through Gemini Enterprise.
It also matters because the partnership explicitly names Agent-to-Agent, Agent-to-UI, and Model Context Protocol support. That is a more serious architecture signal than a normal assistant integration. It suggests the companies are preparing for a world where HR, finance, analytics, and workflow agents pass context and tasks between each other inside one broader enterprise flow.
The business impact lands in operating loops, not chatbot UX
For business teams, the most important shift is that HR and finance support moves closer to repeatable operating loops.
Employee self-service becomes a workflow surface
Questions about leave, payroll, tax withholding, expense policies, or employee records often create a long trail of portal clicks, tickets, and context switching. Putting Sana inside Gemini Enterprise reduces the number of surfaces an employee has to touch, while still routing actions through Workday’s rules and approvals.
Finance work moves from reporting toward action
Workday said the deeper connection between Workday Data Cloud and Google Cloud can move teams from static reports to immediate action. The zero-copy approach is especially notable for finance and HR because it keeps data in place while still allowing query and analysis. That is a more credible enterprise pattern than exporting sensitive operational data into disconnected AI tooling.
Governance becomes part of the interface decision
Workday launched its Agent System of Record in February 2025 as a way to govern fleets of first-party and third-party agents. The May 28 partnership matters because it extends that control-plane idea into the interface layer. The real enterprise question is no longer only which model is best. It is also which surface an employee uses, where approvals happen, how agents hand off work, and whether data stays inside trusted business systems while all of that happens.
What changed in the buying conversation, and what to watch next
This story looks bigger now because the buyer conversation is moving from “Should we use AI in HR or finance?” to “Where should these agents live, and who governs the handoffs?” Workday and Google Cloud are answering that with a combination of workplace-surface access, model standardization, interoperability, and data controls.
- Availability is still staged. Sana Self-Service Agent in Gemini Enterprise is available in early access for eligible Workday customers, not as a broad default rollout.
- More Workday agents are coming later in 2026. Workday said additional agents will arrive later this year, which will show whether the current self-service story expands into deeper role-based finance and HR execution.
- Workday Data Cloud is still early. Workday said the product is available to early adopters now and planned for general availability later this year, so the data-layer part of the story still needs a broader market test.
- Systems integrators are part of the product story. Accenture, Deloitte, and KPMG were named as deployment partners, which is a reminder that large enterprises still see agent rollout as an operating-model and process-change problem, not only a software feature decision.
- Alphabet is also using the stack internally. Workday said Alphabet will use Gemini Enterprise Agent Platform to build a custom Workday agent for Workday administrator workflows, giving the partnership an internal enterprise use case to watch.
The missed-news takeaway is straightforward: Workday and Google Cloud did not just announce another AI partnership on May 28. They made a sharper bet that enterprise agents will win when they can operate inside the employee tools people already use, while policies, approvals, and sensitive transactions stay anchored in the business system of record. That is why this release still has search value now, and why HR and finance leaders should keep watching what Workday ships next.