Workday and Anthropic announced on October 13, 2026, a new partnership that connects their platforms and artificial intelligence agent experiences in human resources and finance. The announcement came at Workday Rising 2026, the company’s annual conference in Las Vegas. The core idea is to combine the reasoning of Claude, Anthropic’s model, with Sana’s agents and with the data and business rules that thousands of organizations already manage in Workday.
Sana is Workday’s agentic experience and orchestration layer, stemming from the Swedish AI company Workday acquired. With the new partnership, work can start in Sana or in Claude and continue on the other platform without losing context or controls. Both exchange information and tasks through the Model Context Protocol (MCP), the open standard Anthropic championed for connecting AI models to external tools and data, so work “carries forward rather than starting over,” according to the announcement.
In Sana, the platform orchestrates Workday’s agents through complex HR and finance processes, such as a compensation cycle or the financial close, which used to take weeks and which the company wants to turn into continuous background processes. Sana holds the context of the process from start to finish and builds the workspace each task needs — for example, a live compensation grid or an exceptions queue — without moving data out of Workday.
When a specific step calls for it, Sana draws on Claude’s reasoning and shares only the context that step needs. The example the companies give is drafting the note a manager uses to explain an employee’s raise. Exceptions go to a person for review. This “minimum necessary context” design aims to limit exposure of sensitive data, a critical point when salaries, reviews or financial information are involved.
From Claude, employees will be able to access Sana agents, analyze Workday HR and finance data and take action without switching tools. A hiring manager could compare open roles against existing headcount and budget and build an internal mobility plan that fills those roles with people already on the team. A finance leader could check invoice exceptions against purchase orders and approve them in Claude or route them to Sana.
The companies stress that every action will follow the customer’s existing permissions and audit controls in Workday, and that Claude will surface only what each person is permitted to see. The user also confirms every change. Admins connect the system once, with no new permission model to build and no per-user setup. For customers, that continuity in governance is probably the partnership’s most important selling point.
Joint customers will also be able to build their own agents directly in Claude using Workday Agent-Ready Tools, with the same governance, security and audit trail built in from the start. Those tools are already generally available to all customers through Workday Extend AI and will come to the plugin later this year. The Sana in Claude plugin, for its part, is available to early access customers and is expected to be generally available in late 2026.
“Customers want the best domain AI agents for both their people and money that understand the business rules they run on and leverage the extraordinary reasoning in Claude,” said Gerrit Kazmaier, president of product and technology at Workday. Ankur Sinha, head of engineering for enterprise and verticals at Anthropic, noted that HR and finance decisions shape every employee’s pay, role and career, and that the combination lets customers plan a compensation cycle or close the books with their own controls applied at every step.
The relationship between the two companies is not new. Anthropic is a Workday customer and runs its finance and people operations on Workday Financials and Workday Human Capital Management, subject to the same rules, approvals and audit controls. Conversely, Workday has rolled out Claude Enterprise, plans to start using Cowork and uses Claude Code in its product and development teams to build its own products. Workday Developer Agent also connects to Claude to help developers.
The announcement reflects a clear trend in the enterprise market: business software vendors want their agents to coexist with the general-purpose AI assistants employees already use, rather than compete with them. For Workday, which says it has more than 11,500 customer organizations, including more than 65% of the Fortune 500, the key is that the rules stay anchored in its platform. For Anthropic, the partnership opens the door to critical processes at large companies.
The bet on MCP is also significant. The protocol has become a common way for agents and applications to exchange data and actions, and adoption by a vendor the size of Workday strengthens its role as a de facto standard. At the same time, security researchers have documented vulnerabilities in MCP implementations, which forces vendors to take care with authentication, permissions and validation of every action, something Workday and Anthropic say they have prioritized.
Some details remain to be settled: pricing, licensing requirements for joint customers and the exact availability timeline by region. It will also be important to see how these agents perform in real processes with imperfect data and frequent exceptions, where human confirmation of every change can be both a safeguard and a bottleneck. Workday scheduled sessions on the partnership at its conference on October 13 itself.
Overall, the Workday–Anthropic partnership points to a model in which a frontier model’s reasoning is combined with the rules, permissions and audits of an enterprise system of record. If it works as promised, employees could handle HR and finance tasks from the tool they prefer, without the company giving up control over who sees what and who approves each change.
