Dreamforce 2026 concluded after three days of major announcements focused on AI’s growing role in business. Held in San Francisco from September 15–17, Salesforce unveiled AIforce, Koa, Agentforce updates, Headless 360, Data 360, Tableau Next, AgentExchange, Trusted Enterprise AI, and expanded technology partnerships.
More importantly, the event showed Salesforce shifting beyond the traditional CRM interface by making data, workflows, business logic, permissions, and actions available across AI interfaces, agents, applications, and digital experiences, creating more connected and governed business workflows.
What Was the Biggest Theme of Dreamforce 2026?
Dreamforce 2026 showed Salesforce converting traditional CRM into the Agentic Enterprise. With this, people and AI agents can work with the same trusted business context. Salesforce is bringing together important applications, workflows and governance to help organizations move from AI assistance to AI-driven action. This shift enables businesses to:
- Give AI business context: Connect with agents and also with trusted data, customer information and required knowledge.
- Move from tasks to actions: This advancement allows agents to reason through processes to take authorized actions in a targeted context.
- Connect work across interfaces: Designed to bring Salesforce capabilities into the tools and AI interfaces where employees already work.
- Combine people and agents: Allow employees to guide, collaborate with and oversee AI throughout business processes.
- Build with trust and control: Apply enhanced security to existing permissions, business rules, and governance for agent-driven work.
Dreamforce 2026 Announcements: Biggest Salesforce Updates
1. AIforce
AIforce works as a new interface layer that helps to bring Salesforce data, business logic, and permissions together into different AI interfaces and work environments. Customers can use Salesforce capabilities without always working through the traditional Salesforce interface.
Key capabilities include:
- Trusted Salesforce context: Brings customer data, business knowledge, workflows, and logic into AI interfaces.
- Business actions: Allows connected AI experiences to work with Salesforce processes and actions.
- Existing controls: Carries Salesforce permissions, security, and governance into connected experiences.
- Multiple interfaces: Launches with Claudeforce, Slackforce, and Agentforce Coworker.
- Developer access: Uses MCP servers, APIs, skills, and plug-ins to create customized AI experiences.
Customer impact: Businesses can extend their existing Salesforce data, workflows, and controls into new AI interfaces without rebuilding the same business logic for every environment.
2. Agentforce
Salesforce expanded Agentforce with a portfolio of job-ready AI agents designed for specific business functions. Customers can start with agents that already have relevant skills, actions, and data models instead of building every agent from scratch.
Key agents include:
- Hunter: Handles outbound sales activities and pipeline work.
- Casey: Supports customer service across voice, SMS, WhatsApp, and web chat.
- Fin: Handles customer operations and service workflows.
- Paige: Supports IT and HR employee requests.
- Carter: Assists shoppers with product discovery and purchases.
- Marshall: Handles supply chain and back-office processes.
- Piper: Engages and qualifies inbound leads.
These agents connect with Customer 360 and can be adapted to a company’s existing processes, data, and business rules. Salesforce also introduced Agent Script to give developers more control over agent behavior.
3. Koa
Salesforce and NVIDIA introduced Koa, Salesforce’s first CRM reasoning model for Agentforce. It is built on NVIDIA Nemotron and designed specifically for complex, multi-step CRM workflows.
Key capabilities include:
- CRM-specific reasoning: Designed for Salesforce sales and service scenarios.
- Multi-step problem solving: Helps agents work through complex business processes.
- Tool selection: Helps agents determine which tools and actions are needed.
- Enterprise context: Uses Salesforce CRM knowledge to improve agent reasoning.
- NVIDIA collaboration: Built through Salesforce and NVIDIA technical collaboration using NVIDIA Nemotron.
Salesforce says Koa was developed using synthetic enterprise data modeled on knowledge from nearly three decades of CRM deployments and was not trained on customer data.
Customer impact: Agentforce can use a reasoning model specifically designed for CRM work rather than relying only on general-purpose model capabilities.
4. Agentforce Coworker
Agentforce Coworker brings an autonomous AI assistant directly into employee workflows. It can answer questions, reason through tasks, orchestrate work, and operate within the permissions of the logged-in employee.
Key capabilities include:
- Employee assistance: Helps users find information and complete work.
- Task orchestration: Can reason through multiple steps instead of only answering questions.
- Permission-aware actions: Works within the user’s existing access boundaries.
- AI Skills: Employees can teach reusable skills for specific tasks.
- Cross-interface work: Can operate across Salesforce and connected work environments.
Salesforce also highlighted the Adecco Group’s rollout of Agentforce Coworker across 40 countries.
Customer impact: Employees can delegate routine or multi-step work to an AI coworker while maintaining existing access and governance controls.
5. Headless 360
Headless 360 turns Salesforce capabilities into reusable services that authorized AI agents, applications, and digital experiences can access through open standards. It extends Salesforce beyond its traditional application interface.
Key capabilities include:
- Headless 360 MCP Server: Allows AI agents to discover and use Salesforce capabilities through MCP.
- Data 360 MCP Server: Extends Data 360 context to external AI environments.
- Agent Skills and Plugins: Packages business logic and processes into reusable capabilities.
- Headless Experience Layer: Supports web, mobile, embedded, and conversational experiences.
- Salesforce Multi-Framework: Supports modern React applications with Salesforce authentication, security, and data access.
- Expanded cloud support: Extends headless capabilities across Sales, Service, Marketing, Commerce, MuleSoft, Informatica, and Tableau.
Customer impact: Businesses can reuse existing Salesforce business logic across AI agents, websites, applications, and other interfaces instead of rebuilding integrations for each new experience.
6. Data 360: Building the Data Foundation for AI
Data 360 remains a core part of Salesforce’s Agentic Enterprise strategy because AI agents need unified, governed business data and context to make useful decisions.
Key capabilities include:
- Unified customer context: Combines data from different sources into a connected customer view.
- Identity resolution: Helps match customer records across systems.
- Semantic context: Gives AI a greater understanding of what business data means.
- Data 360 MCP Server: Exposes Data 360 capabilities to authorized AI agents.
- Zero-copy connectivity: Allows organizations to work with data across external systems without always moving or duplicating it.
- AI-ready data operations: Agents can support activities such as data transformation, calculated insights, audience creation, and activation.
Customer impact: Customers can provide agents with a more complete and governed business context instead of relying on fragmented CRM records.
7. Tableau Next
Tableau Next expands Salesforce analytics toward an agentic model where AI can help users understand data, identify important signals, and take action from those insights.
Key capabilities include:
- AI-powered insights: Surfaces important patterns and signals from business data.
- Semantic models: Connects analytics to define business meaning.
- Predictive analytics: Supports forecasting and identifying key business drivers.
- Conversational analysis: Lets users explore data using natural language.
- Action from insights: Connects findings with Salesforce workflows and business processes.
- Agent integration: Brings Tableau analytics into agentic workflows.
Tableau Next works with Data 360 and Salesforce context to connect analytics with business actions.
Customer impact: Teams can move from finding information in dashboards to using AI-supported insights as part of everyday business decisions and workflows.
8. AgentExchange
AgentExchange gives businesses a marketplace and ecosystem for discovering agents, skills, apps, integrations, and other components that can be used with Salesforce’s agentic platform.
Key capabilities include:
- AI agents: Prebuilt agents for different business functions.
- Agent Skills: Reusable capabilities that agents can perform.
- MCP servers: Connect AI agents with enterprise capabilities.
- Apps and integrations: Extend agents across business systems.
- Partner solutions: Allow technology partners to build on Salesforce’s agentic architecture.
- Headless capabilities: Support experiences outside the traditional Salesforce interface.
Salesforce says AgentExchange works with the broader Headless 360 architecture, allowing partners to build interfaces, agents, applications, integrations, workflows, and actions.
Customer impact: Businesses can adopt existing agent capabilities and integrations rather than developing every AI function internally.
9. Trusted Enterprise AI
As AI agents gain access to business data and the ability to take action, Salesforce is adding a broader architecture for controlling how AI operates across an organization.
Key capabilities include:
- Trusted Context: Gives AI governed business data, knowledge, metadata, and memory.
- Trusted Agency: Supports reasoning, planning, memory, and orchestration.
- Trusted Action: Connects agents with applications, APIs, workflows, and business processes.
- Trusted Governance: Applies policies, controls, data quality, and lineage.
- Trusted Security: Controls identity, permissions, privacy, and data access.
- Trusted Models: Allows organizations to connect different AI models based on business requirements.
- AI Control Plane: Provides a central place to manage agents, policies, performance, lifecycle, and costs.
Customer impact: Organizations get a centralized approach to controlling what AI can access, what it can do, which models it uses, and how its actions are governed.
10. Expanded AI Partnerships
Salesforce expanded its AI partnerships to connect Salesforce data, agents, models, and infrastructure with other major technology platforms.
Key developments include:
- Anthropic: Claudeforce connects Claude with Salesforce data, workflows, and business logic.
- NVIDIA: Koa brings NVIDIA Nemotron into Salesforce’s CRM reasoning capabilities.
- Google Cloud: Salesforce and Google Cloud expanded their partnership so agents across both platforms can work with shared data and infrastructure.
- AWS: Salesforce expanded its collaboration with AWS around CRM data, AI agents, infrastructure, and model choice.
- OpenAI: Salesforce expanded its relationship with OpenAI through Missionforce and government-focused AI capabilities.
The Google Cloud partnership also connects Salesforce’s headless architecture with Gemini Enterprise, allowing agents on both platforms to reason and act on shared data.
Customer impact: Customers can connect Salesforce with major AI models and cloud platforms while keeping Salesforce data, workflows, permissions, and business logic at the center.
11. Real-World Customer and Industry Deployments
Dreamforce 2026 also showed how Salesforce customers are applying Agentforce across sales, service, employee support, government, entertainment, and industrial operations.
Key examples include:
- Siemens: Uses it to engage inbound leads and connect sales processes with industrial customer information.
- TSA: Introduced Ace, an AI agent designed to answer traveler questions and support airport security information.
- Live Nation: Uses Agentforce to provide customer support for fans across U.S. venues.
- Adecco Group: Is rolling out Agentforce Coworker across 40 countries.
- Engine: Uses Salesforce capabilities to power its EVA AI support agent.
- Hibbett, Perk and Autism Queensland: Featured as examples of organizations applying job-ready capabilities.
These examples cover different types of work, from customer conversations and lead engagement to employee support and industry-specific processes.
Customer impact: For real understanding of agentic technology with real business environments Salesforce provides practical examples to customers beyond just product demonstrations.
Benefits of Dreamforce 2026 for Salesforce Customers
At Dreamforce, Salesforce announced to give customers new ways to connect AI agents that help deliver advanced automated workflows. The benefits extend beyond automation to provide organizations with improved service, connected systems, and maintain greater control over AI-driven work.
- AI agents can handle routine customer requests, provide responses across channels, and support service teams with relevant customer context.
- Data 360, AIforce, and conversational analytics can make relevant business information available through the interfaces employees already use.
- Headless 360, APIs, MCP servers, and Agentforce can connect Salesforce capabilities with external applications and business processes.
- Tableau Next connects business data with AI-supported insights, helping teams move from analyzing information to taking action.
- Customers can use different AI models, agents, applications, and interfaces based on their business requirements.
- Agent Skills, MCP servers, APIs, and AgentExchange allow organizations to reuse existing capabilities across different agents and experiences.
What Dreamforce 2026 Means for Salesforce Professionals
This year Dreamforce changes how Salesforce professionals will work with AI, automation, data, and agents. Each role will have new responsibilities as Salesforce moves toward the Agentic Enterprise.
Impact on Salesforce Administrators
Admins will spend less time on repetitive tasks and more time managing AI agents, permissions, workflows, data, and governance.
Impact on Salesforce Developers
Developers will build AI agents, Agent Skills, integrations, APIs, and headless experiences while working with tools such as MCP.
Impact on Salesforce Consultants
Consultants will help businesses identify where AI agents can add value, design workflows, and connect Salesforce with existing systems.
Impact on Salesforce Architects
Architects will focus on AI architecture, data strategy, security, integrations, model selection, and how agents work across the Salesforce ecosystem.
What Salesforce Customers Should Evaluate After Dreamforce 2026
The AI capabilities announced at Dreamforce 2026 don’t necessarily mean that every organization must implement all of these capabilities at once. The more realistic way to do this is to assess the role of agentic technology in current business processes, data architecture and governance.
AI agents are not immune to challenges and hype, so it is important for Salesforce customers to weigh the following before implementation:
Data Readiness
Reliable business context is essential to the operation of AI agents. Before agents are permitted to use business information to make decisions or take action, organizations need to (re)consider data quality, data duplication, identity resolution, metadata and access controls.
Process Readiness
An ambiguously defined business process can be hard to automate in a secure way. If assigning parts of a workflow to an AI agent, teams should document decisions, exceptions, approvals, and dependencies in a workflow prior to deployment.
Permission and security controls
AI agents might have the ability to access information and carry out actions on users’ behalf. Organisations must then decide what data and actions each agent is granted access to as well as whether the permissions and governance provided in Salesforce are adequate for their use case.
Human Oversight
It is not necessary for every process to be completely independent. Businesses can figure out where AI can help complete an action, where human approval is needed, and where AI can do it without being risky for a business agent.
Integration Requirements
Organizations should measure the interaction of agents with current Salesforce applications, APIs, external systems, platforms, and processes. These possibilities can be extended with headless and MCP-based capabilities, but they must be integrated into the organization’s entire integration architecture.
Product Availability
Any presentations about products during Dreamforce should also be kept from what is currently available on the market. For certain products, regions, and customer agreements, Salesforce says the availability of its new capability may differ, and organizations should confirm its availability, licensing, and deployment requirements before adding a new capability to production plans.
The real question is not just whether a business should use AI agents, but which ones.So, the key question following Dreamforce 2026 is, not if, but which AI agents should a business use? It is where agents can make value and what context the business should have, what they can do, and what controls should be kept.
The End Words
So much of Dreamforce 2026 has been focused on Salesforce’s ability to become more than just a CRM data repository. These capabilities, along with Salesforce’s ability to connect the business data, permissions, workflows and governance, form a model that allows Salesforce to be accessible to users, apps and AI agents. Melonleaf Consulting’s move underscores the critical role of linking Salesforce functionalities to real-world business processes powered by AI.
The next step for Salesforce customers will then be more about determining the role AI plays within current business processes, rather than just integrating AI into the system. In this era of AI support, the importance of AI-driven processes and data quality, process design, security, and integration architecture, along with human oversight, will remain key.
