Workday's 'Agent System of Record' Hits General Availability — HR Now Manages AI Workers

Workday confirmed this week that its Agent System of Record (ASOR) — first announced in February 2025 — is now generally available across all major Workday tiers. The platform brings AI agents into the same operating layer where HR teams manage employees: onboarding, role definition, cost tracking, performance metrics, compliance audit trail. More than 65 global partners and nearly 20 Workday Ventures portfolio companies have already connected agents to the system. The framing — that AI agents need to be 'hired,' supervised, budgeted, and offboarded like any other worker — is the most consequential part.

Why a 'System of Record' for AI Agents Matters

Through 2025, most enterprise AI agent deployments lived as one-off integrations: a customer-service bot from one vendor, a finance-automation agent from another, a sales-research agent built in-house. Each was managed in its own console with no unified visibility into which agents had access to which data, what they were costing, or how their outputs were being audited. ASOR is Workday's bet that the answer to that fragmentation is the same answer enterprises adopted for human employees thirty years ago: a single system of record. Define the agent's role and reporting line, gate its data access, attach a budget, attach an owner, attach a compliance checklist.

The New Job Titles This Creates

Workday's go-to-market materials describe roles like 'Agent Supervisor,' 'Agent QA Lead,' and 'AI Operations Manager' as the people who will actually use ASOR day-to-day. These titles are no longer hypothetical — Workday cites customer rollouts where companies have created formal Agent Operations functions reporting either into IT, HR, or a newly-created Chief AI Officer org. The salary bands referenced in early 2026 customer pilots cluster between $130K and $250K depending on scope, with the supervisor tier typically requiring three years of operations or program-management experience plus demonstrated AI tooling fluency rather than ML engineering credentials.

What HR Teams Actually Have to Do Now

For HR leaders, ASOR's general availability shifts AI from an IT problem to a workforce-management problem with quarterly KPIs attached. Agent budgeting moves into the same forecast cycle as headcount. Agent compliance — the question of which AI was authorized to take which action against which dataset — moves into the same audit framework as employment compliance. The skill profile of HR teams is shifting fast as a result: the EVP-of-People in 2026 increasingly needs functional fluency in agent governance, vendor evaluation, and AI cost modeling alongside traditional comp, benefits, and culture work.

The Build-vs-Buy Decision Got Harder

ASOR's existence also reframes a strategic decision most enterprises are now actively making: do they build an internal agent governance layer, buy ASOR (or a competitor like Salesforce's Agentforce console or ServiceNow's AI Control Tower), or compose multiple. The pattern emerging from early customer references is that companies already heavily invested in Workday for HCM are defaulting to ASOR by inertia, while companies with Salesforce or ServiceNow as their primary system of record are pulling agent governance into those platforms instead. The losing strategy in either case is no governance layer at all — by 2027 that's a compliance liability, not a cost-savings story.