The 600 AI Agents Inside Oracle Cloud HCM: Module-by-Module Guide

Most online sources mention Oracle’s HCM AI with one sentence: “AI is embedded.” In reality, that’s not helpful, is it?

So, I’ve reviewed the official Oracle Fusion AI document that names every agent by module, release, and type. And combined it with Oracle’s March 2026 announcement of 600-plus embedded agents across Fusion Cloud, the real picture is much bigger.

Therefore, here you can find agent-level details that no website provides. I’ll show you exactly which agents arrived in 26A, which are coming in 26B and 26C, and what each one actually does for your team.

If you manage HRIS, payroll, or consulting for organizations with 500 or more employees, this inventory will help you plan activation and training. Let’s look at what’s really inside.

Key Findings

These facts come straight from official Oracle documentation and verified analyst reports.

Key Finding	Why It Matters for You
600 plus embedded AI agents across Fusion Cloud included at zero extra licensing cost. Announced March 2026.	You don’t pay extra for AI. Workday and SAP typically charge consumption fees.
Every agent is documented by module with exact release versions from 24D through 26C.	Most competitors say “AI embedded.” This is the only public inventory naming agents.
Strategic agents (Jobs Assistant, Workforce Structures Insight Analyst) arrived in 26A. These are planning tools.	You get workforce planning agents that competitors don’t offer natively.
Agent Studio allows custom agent creation with multiple LLM selection at no extra cost.	You can build organization-specific agents without custom development contracts.
Governance is required. Pilot, collect feedback, maintain audit trails. All AI runs inside Oracle’s system.	Adoption requires human review. Oracle provides guardrails, but planning is yours.

Why “AI is Embedded” Misses the Point

Most competitor articles use generic language. They say Oracle “uses AI for recruiting” or “offers embedded intelligence.” That’s not useful when you’re deciding which agents to activate.

I’ve checked every verified agent listed in official Oracle documentation. Most blogs never name a single agent by its actual function. They don’t explain whether it’s an agent, a generative feature, or a predictive model.

This guide fixes that gap. You’ll see the complete module-level inventory. You’ll also see release versions, so you know what’s live now and what’s coming next.

Well, let’s look at what “embedded” actually means for your daily work.

The 600 Plus Agent Ecosystem: What It Actually Means

Oracle announced 600 plus pre-built AI agents embedded across Fusion Cloud in March 2026. You may have seen that headline. What’s less clear is how many apply to HCM specifically.

The official documentation (updated June 10, 2026) lists agents across five HCM modules. Each entry includes the release version, the type (Agent, Generative, or Predictive), and the exact business action.

Here’s the critical distinction you need to know. A copilot answers questions. It drafts content. But an agent executes transactions. It checks policies. It reports outcomes within guardrails. That’s a big difference.

So let’s look at what’s actually available.

Agent Types Explained (Before the Inventory)

Before I show you the full list, here’s a quick breakdown of the three types you’ll see.

• Agent means it takes action. It validates data, generates profiles, or advises based on policies. You configure guardrails.
• Generative means it creates text. It writes instructions, descriptions, or assessments. You review before using.
• Predictive means it forecasts or suggests. It infers skills or predicts turnover risk based on patterns.

You can configure audit trails and escalation rules for agents. You can review generative output before approval. Predictive results feed into reports.

Now let’s see the inventory.

Complete Agent Inventory by Module (Verified from Official Docs)

I’ve organized this by module so you can scan. Each agent includes the release version, type, and the concrete business action.

Here are the verified agents from the official readiness documentation.

HCM Common Agents

These agents support data management, employee self-service, and team management across all HCM functions.

The AI Agent for HCM Data Loader arrived in 26A. It validates bulk imports. It suggests corrections based on previous loads. It flags conflicts against security rules.

You may find this valuable during acquisitions or large hiring events. Before this agent, you needed manual validation for high-volume imports.

The Employee Self-Service agent received enhanced AI assistance in 26A. It’s using company policy documents to deliver tailored support. When an employee asks about direct deposit changes, it retrieves the policy and guides them through the form.

The Team Management agent also received 26A enhancements. It’s integrating workforce data to provide recommendations for managers handling team-level actions.

There’s also a Spreadsheet Data Loader Advisor coming in 26C. It advises on spreadsheet-based bulk loading workflows.

Here’s the HCM Common table.

Agent / Feature	Release	Type	What It Does for You
AI Agent for HCM Data Loader	26A	Agent	Validates bulk imports; suggests corrections
Employee Self-Service with AI Assistance	26A	Agent	Policy-aware employee guidance
Team Management with AI Assistance	25D, enhanced 26A	Agent	Manager-level team recommendations
Spreadsheet Data Loader Advisor	26C	Agent	Advises on spreadsheet bulk loading
Resolve HCM Data Security Access	26C	Agent	Helps fix user access and security

Human Resources Agents

The Jobs Assistant arrived in 26A. You can give it a job name and job code. It generates a complete job profile using organizational rules, skill requirements, and existing templates.

I’m seeing this reduce manual profile creation significantly. Before this agent, you needed manual entry for every new role description.

The Workforce Structures Insight Analyst also arrived in 26A. You can ask it to pull workforce data and apply scenario modeling. You may ask what labor costs look like if you shift 15 percent of full-time staff to part-time in Q3. It presents a structured insight report.

There’s a Talent Profile Search agent coming in 26C. It searches talent profiles by skills for team planning.

Here’s the HR agent inventory.

Agent	Release	Type	Business Action
Jobs Assistant	26A	Agent	Generates job profiles from name/code
Workforce Structures Insight Analyst	26A	Agent	Scenario modeling and workforce reporting
Intelligent Talent Profile Search	26C	Agent	Searches talent profiles by skills

Compensation Agents

The Compensation Plan Instruction Refinement feature arrived in 26A. Before 26A, you could generate instructions from scratch. Now you can refine them directly in the workflow.

Admins have five AI options: revise, expand, summarize, formalize, or edit. You have the option to apply any of these to AI-generated instructions before final approval.

The Peer-Group Salary Benchmarking feature also arrived in 26A. Managers can compare team salaries against peer-group market composites. You may use this before proposing salary adjustments.

Here’s the compensation timeline.

Agent / Feature	Release	Type	What It Does for You
Compensation Plan Instruction Refinement	26A	Generative	Refines compensation instructions (5 options)
Peer-Group Salary Benchmarking	26A	Agent	Peer-group salary comparison
Compensation Advisor (My Team)	25C, enhanced 26B	Agent	Team compensation advice with market data
Line Manager Salary Change Agent	26B	Agent	Validates manager-proposed salary changes
Workforce Compensation Analyst	26C	Agent	Workforce-level compensation analysis

Absence Management Agents

These agents support absence tracking, policy guidance, and time-off management.

• The Leave and Absence Analyst arrived in 25C. It analyzes team absence data and trends. Managers can review patterns before scheduling.
• The Leave and Absence Policy Advisor from 25C continues to advise employees and managers on absence rules. It uses uploaded company policy documents.
• The Time-off Assistant in 26A received enhanced policy support. When an employee asks for time off, the agent retrieves the current policy document. It guides the user through the correct form.
• There’s also a Policy Document Upload feature in 26A that enables these agents to access company absence policies directly.

Here’s the absence agent table.

Agent	Release	Type	Business Action
Leave and Absence Analyst	25C	Agent	Team absence data analysis
Leave and Absence Policy Advisor	25C	Agent	Policy-based absence guidance
Time-off Assistant	26A	Agent	Policy-aware time-off assistance
Leave and Absence Analyst (My Team)	25C	Agent	Manager-level absence insights
Policy Document Upload (Agent)	26A	Agent	Enables policy retrieval for agents

Benefits Agents

Benefits agents support enrollment guidance, policy advice, and certification management.

• The Benefits Analyst arrived in 25C. It provides guidance to employees and managers on benefit questions.
• The Benefits Policy Advisor from 25C advises on benefit policies using company documents. You may configure which policies it retrieves.
• The Benefits Plan Advisor Agent from 25D advises employees on benefit plan selection. It’s useful during open enrollment.
• Pipeline agents include a Court Order Intake Assistant (26B) and a Benefits Certification Agent (26B).

Here’s the benefits table.

Agent	Release	Type	Business Action
Benefits Analyst	25C	Agent	Benefits guidance for employees/managers
Benefits Policy Advisor	25C	Agent	Policy-aware benefits advice
Benefits Plan Advisor	25D	Agent	Benefit plan selection guidance
Court Order Intake Assistant	26B	Agent	Assists with court-ordered adjustments
Benefits Certification Agent	26B	Agent	Manages benefit certifications

Dynamic Skills Agents

Dynamic skills agents manage skill libraries, assessments, and learning recommendations.

The base Dynamic Skills predictive model arrived in the August 22C maintenance pack. It infers skills from employee profiles.

The Generate Skill Description agent from 25C creates descriptions from library templates. The Multi-Language Skill Description Agent from 25D extends this to multiple languages.

In 26A, the Skill Suggestions for Learning Items feature uses generative AI. It suggests relevant skills based on learning content descriptions.

The Intelligent Talent Profile Search agent arrives in 26C. You can search talent profiles by skills for team planning and succession.

Here’s the skills timeline.

Feature / Agent	Release	Type	What It Does for You
Dynamic Skills (Base Predictive)	Aug 22C	Predictive	Predictive skills inference
Generate Skill Description	25C	Generative	Creates skill descriptions
Skill Description Agent (Multi-Language)	25D	Generative	Multi-language descriptions
Skill Rating Assessments	25C	Generative	Creates skill assessments
Skill Suggestions for Learning Items	26A	Generative	Suggests skills for learning content
Intelligent Talent Profile Search	26C	Agent	Searches talent profiles by skills

Real-World Use Cases: What These Agents Actually Do

Below are concrete scenarios. I designed these based on the verified agent actions from official documentation.

These aren’t theoretical examples. Each one matches a verified agent action.

Strategic Planning: Jobs Assistant + Workforce Structures Insight Analyst

A manufacturing company needs 150 new supervisor roles. Before 26A, HR managers wrote profiles manually.

The Jobs Assistant generates complete profiles using organizational rules and templates. You provide the job name and code. It handles the rest.

Then the Workforce Structures Insight Analyst pulls workforce data. It models labor costs for Q2 versus Q3 hiring. You get a structured insight report.

This isn’t automation for automation’s sake. It’s strategic planning with structured data.

Daily Operations: HCM Data Loader + Self-Service Agent

A retail company acquires 2,000 employees. Before 26A, bulk imports required manual validation.

The Data Loader Agent validates mappings. It flags conflicts. It suggests corrections based on previous loads.

New hires ask the Self-Service Agent about direct deposit updates. It retrieves the company policy. It guides them through the correct form.

Your team spends less time on manual corrections. Employees get faster answers.

Compensation Planning: Instruction Refinement + Benchmarking

A compensation team revises plan instructions for a new equity program. Before 26A, they wrote instructions from scratch.

Now they use the Refinement Agent. They have five options: revise, expand, summarize, formalize, or edit. They pick the right option for their audience.

Managers then use the Peer-Group Benchmarking Agent. They compare team salaries against market composites before proposing adjustments.

The result is consistent communication and data-backed manager decisions.

Absence Management: Policy Advisors + Time-Off Assistant

An employee asks for extended leave. Before 26A, they might call HR for policy clarification.

The Time-off Assistant retrieves the current absence policy. It guides the employee through the request form.

The Policy Advisor supports managers reviewing the request. It checks entitlement formulas against company rules.

Your HR team sees fewer routine policy questions. Managers handle approvals faster.

Release Timeline: When Each Agent Arrived

This timeline connects previous releases to future pipeline. You may use it to plan training and activation schedules.

Here is the verified timeline.

Agent / Feature	Release	Strategic Importance
Dynamic Skills (Base Predictive)	Aug 22C (Maint. Pack)	Foundation for skills inference
Compensation Advisor (My Team)	25C	Team compensation guidance
Benefits Analyst / Policy Advisor	25C	Benefits support
Leave and Absence Agents	25C	Absence automation
Dynamic Skills Description / Ratings	25D	Multi-language skills
Role Guides / Enhanced Filtering	25A	Navigation and onboarding
Total Compensation Statement Agent	25A	Transparency
Jobs Assistant / Workforce Structures	26A	Strategic workforce planning
HCM Data Loader / Self-Service / Team Mgmt	26A / 25D	Data and daily operations
Redwood Transaction Console AI	26B	Embedded generative AI in transactions
Enhanced Time-off / Line Manager Salary Agent	26B	Enhanced absence and compensation control
Spreadsheet Data Loader / Compensation Analyst / Dynamic Skills Search	26C	Advanced analysis and strategic planning

Building Custom Agents: Agent Studio for HCM

Almost no competitor content covers this. Yet Oracle includes Agent Studio at zero extra licensing cost.

You have the option to build custom agents for organization-specific HR processes. You can select from multiple LLMs. You can deploy partner templates.

Before building, you’ll want to follow a governance framework. Pilot in one HR function. Collect feedback. Monitor accuracy.

Here’s what Agent Studio requires.

1. Define your process. Pick one HR workflow (onboarding, absence approval, compensation review).
2. Configure policy guardrails. Set rules for what the agent can and cannot approve.
3. Select your LLM. Oracle supports multiple LLM options within Studio.
4. Deploy a pilot. Test with a small user group before full rollout.
5. Monitor and iterate. Track adoption, accuracy, and user satisfaction.
6. Document configurations. Maintain audit trails for compliance.

Competitive AI Licensing: Oracle vs. Workday vs. SAP

Here’s the comparison buyers need to see. Most articles skip this entirely.

Oracle’s 600 plus embedded agents come at zero additional licensing cost. Workday and SAP generally charge extra through consumption-based or module-based pricing.

Here’s the strategic comparison table.

Dimension	Oracle 26A / Fusion AI	Workday	SAP SuccessFactors
Agent count (verified public docs)	600+ embedded agents (official docs list module-level agents)	Not disclosed as embedded count; AI requires additional licensing	Not disclosed as embedded count; AI requires additional modules or fees
Licensing cost for AI capabilities	Zero additional cost (included in subscription)	Typically consumption-based or module fees	Typically module or consumption fees
Agent types documented by module	Agent, Generative, Predictive (official docs list by module and release)	Primarily generative/copilot features	Primarily predictive + generative via partners
Custom agent framework	Agent Studio included (multi-LLM, policy guardrails)	Limited custom framework	Limited custom framework
Strategic planning agents (verified)	Jobs Assistant, Workforce Structures (26A)	Not available as native embedded agents	Not available as native embedded agents
Implementation acceleration	Best-practice configurations built in; faster than SAP	Comparable for core modules	Slower due to customization requirements

Governance, Ethics, and Adoption Best Practices

All agent actions occur inside Oracle’s secure system with audit trails. But adoption still requires planning.

Here’s the framework I recommend based on verified implementation guides.

• Pilot first. Pick one HR function. Recruitment automation works well. So does onboarding assistance.
• Collect feedback. Gather input from HR admins, managers, and employees during the pilot.
• Monitor accuracy. Check that agent outputs match company policies. Adjust guardrails if needed.
• Maintain audit trails. Verify that Oracle’s native audit functionality is enabled for all agent actions.
• Iterate before scaling. Don’t activate all agents at once. Start with HCM Common agents (Data Loader, Self-Service). Then expand to strategic agents.

Source: Details available at Kovaion and Smactworks.

Frequently Asked Questions (FAQ)

These answers address the most common gaps in competitor content. They’re based only on verified sources.

Final Thoughts and Verdict

Overall, I’ve counted every verified agent listed in official Oracle documentation. Most competitor content ignores agent names entirely. This guide fixes that gap.

If you manage Oracle Cloud HCM, you have dozens of verified agents available right now. You have strategic planning agents in 26A that competitors don’t provide natively. And you have custom agent capabilities through Agent Studio at zero extra licensing cost.

The organizations that gain the most from this release will be those that test agents in phases, configure guardrails properly, and train users by role. Don’t activate everything at once. Start with HCM Common agents. Then expand to strategic planning.

If you manage a global workforce with 500 or more employees, the licensing economics alone make Oracle’s embedded AI model worth serious evaluation.

The agent inventory here gives you the only public reference to decide which capabilities to activate first.