AIforce Explained: How Salesforce's New Interface Layer Actually Works
Every Dreamforce produces a headline feature. This year, it was AIforce, and unlike most keynote announcements, it isn't really a feature at all. It's a repositioning of what using Salesforce even means.
For executives evaluating what this means for their organization, the useful exercise isn't memorizing the new product names. It's understanding the architecture underneath them, because that architecture determines what changes for your teams, your admins, and your governance model — and what doesn't.
What Is AIforce?
AIforce is a new interface layer built on top of the Salesforce platform. It allows users to access Salesforce data, workflows, and agent capabilities from the applications they already use every day.
The AIforce portfolio currently includes:
Underpinning these experiences is the Headless Toolkit, a framework that exposes Salesforce data, permissions, workflows, and business logic beyond the traditional Salesforce interface.
What's new?
Historically, users interacted with Salesforce through Lightning pages, dashboards, reports, and application screens. AIforce expands those access points. Teams can now work with Salesforce data and workflows through conversational interfaces while continuing to operate in the environments where they already collaborate and make decisions.
Understanding AIforce Within the Salesforce Architecture
Salesforce now positions its platform as four connected layers.
The data layer responsible for connecting, harmonizing, and managing enterprise data across systems.
The application layer where business processes, CRM functionality, automation, and workflows operate.
The operational layer where AI agents execute tasks and workflows using organizational data and business logic.
The interaction layer that provides access to Salesforce capabilities through conversational experiences.
How AIforce and Agentforce Work Together
Since the announcement, many Salesforce professionals have asked whether AIforce changes the importance of Agentforce.
Salesforce's positioning remains straightforward:
- Agentforce is responsible for execution.
- AIforce is responsible for interaction.
Agentforce continues to power the agent layer, while AIforce provides new ways for users to access those capabilities through Claude, Slack, and Salesforce Lightning.
The Three Ways to Experience AIforce
Claudeforce
Bringing Salesforce Into Claude
Claudeforce is Salesforce's partnership with Anthropic that combines Salesforce data and workflows with Claude's conversational experience.
Current Capabilities
- Prebuilt sales skills currently available in beta
- Salesforce development capabilities through Claude Code
- Support for CRM data and workflow interactions
- Planned expansion into service, marketing, commerce, and industry-specific use cases
What It Looks Like in Practice
Slackforce
Extending Salesforce Into Daily Collaboration
Slack has become a central part of many Salesforce environments, and Slackforce expands the role it plays.
Key Components of Slackforce
Slackforce Surfaces
Slack CRM
Slackbot
Slack Code
Business Impact
Teams can identify issues, collaborate on decisions, act, and update systems without moving between multiple applications. Information entered in Slack remains synchronized with Salesforce, reducing manual re-entry and improving operational efficiency.
Agentforce Coworker
Bringing Conversational AI Into Salesforce
While Claudeforce extends Salesforce into Claude and Slackforce brings it into Slack, Agentforce Coworker delivers the same conversational experience directly inside Salesforce Lightning. Designed for employees who spend most of their day working within Salesforce, Agentforce Coworker provides access to data, workflows, and agent capabilities through natural-language interactions without leaving the CRM environment.
Capabilities of Agentforce Coworker
Context-Aware Assistance
Action Execution
Embedded Salesforce Experience
Unified Access to Business Context
How Does it Work Practically
The Headless Toolkit: The Foundation Behind AIforce
One of the most important architectural developments behind AIforce is the Headless Toolkit.
What is the Headless Toolkit?
What does it do?
What It Enables
- Access Salesforce capabilities through multiple interfaces
- Build custom experiences beyond Salesforce Lightning
- Connect external AI platforms to Salesforce workflows
- Extend Salesforce functionality into third-party applications
Governance and Security Considerations
One of the most common concerns following the AIforce announcement has been governance. As Salesforce expands into multiple interfaces, organizations need assurance that security and compliance controls remain intact.
Existing Controls Continue to Apply
AIforce uses the same:
- Permission sets
- Sharing rules
- Security policies
- Business processes
- Governance frameworks
Users only access information they are already authorized to see.
Beyond these existing controls, Salesforce has introduced three new capabilities purpose-built for agentic governance:
Agentic Identity
Salesforce also introduced Agentic Identity, which tracks and monitors actions performed by both human users and agents. Organizations can define permissions for agents separately and restrict access as needed.
Observability Center
Observability Center provides administrators with visibility into agent activity across the organization, making it easier to monitor behavior and intervene when necessary.
Zero Data Retention
Salesforce has emphasized Zero Data Retention (ZDR) across the AIforce ecosystem. According to Salesforce, business data processed through supported interfaces is not retained or used to train underlying models. For regulated and security-conscious organizations, this will be an important area to evaluate as AIforce adoption grows.
AgentExchange and the Expanding Ecosystem
AIforce is not intended to operate as a closed Salesforce-only environment.
The former AppExchange has been rebranded as AgentExchange, creating a marketplace for:
- Agents
- Connectors
- Integrations
- Industry-specific solutions
- Headless Toolkit extensions
Organizations can leverage solutions developed by Salesforce partners, while partners can create capabilities that extend Salesforce into additional business processes and platforms. The ecosystem includes AI providers, development platforms, workflow solutions, and business application vendors.
What Executives Should Focus On
The introduction of AIforce changes where employees interact with Salesforce. Instead of requiring users to work exclusively inside CRM screens, Salesforce capabilities can now appear inside the applications where employees already spend their time.
Potential Benefits
- Reduced application switching
- Faster access to business information
- Improved workflow efficiency
- Broader adoption of Salesforce data and processes
Areas That Require Attention
Success with AIforce will depend on foundations that already exist within the organization:
- Data quality
- Security architecture
- Permission design
- Agent governance
- Workflow design
- Business process controls
Organizations that have not addressed these areas may find that existing complexity becomes harder to manage when Salesforce is extended across multiple interfaces.
Where Rialtes Fits In
AIforce represents a shift in how users interact with Salesforce. Data, workflows, permissions, and agent capabilities remain within the Salesforce platform, but access to those capabilities is no longer tied to a single interface. It introduces new ways to access Salesforce, but it does not reduce the importance of a strong Salesforce foundation.
Rialtes helps organizations design and implement:
- Data 360 architectures
- Customer 360 environments
- Agentforce deployments
- Governance frameworks
- Security and permission models
- Enterprise integrations
By establishing the right foundation first, organizations can extend Salesforce into Claude, Slack, and other AIforce surfaces while maintaining the governance, security, and operational controls required for long-term success.
AIforce FAQs
Latest Blogs
