Salesforce’s Dreamforce 2026, held in San Francisco from 15–17 September, has arguably changed the way organisations should think about their CRM.
While Agentforce remained a major focus throughout the three-day event, the biggest announcement was arguably not another AI agent. It was AIforce — Salesforce’s new live interface layer designed to bring the full power of the Salesforce platform into the AI interfaces where people and agents already work.
The proposition is simple:
The full power of Salesforce. Any AI interface.
With AIforce, Salesforce is taking the data, workflows, business logic, semantics, permissions, security, and governance already embedded within the Salesforce platform and making them accessible through interfaces such as Claude, Slack, and Salesforce Lightning.
Instead of requiring employees to open Salesforce, navigate dashboards, and search through records to complete a task, Salesforce is moving towards a model where Salesforce comes to the user.
That shift could have significant implications for how businesses approach CRM, AI adoption, automation, and Agentforce implementation.
Introducing AIforce: The Full Power of Salesforce
Much of an organisation’s operational knowledge already sits inside Salesforce.
Customer records, sales activity, service cases, workflows, business rules, permissions, and other critical information have traditionally been accessed through the Salesforce interface.
AIforce changes that model.
Salesforce describes AIforce as the live interface layer for the Agentic Enterprise, bringing Salesforce’s trusted business context into the AI interfaces where people and agents work. Employees can ask questions, update records, and trigger workflows without necessarily opening Salesforce itself.
For example, instead of navigating through multiple Salesforce screens to understand an account, a sales representative could interact with Salesforce through an AI interface and ask for the information they need.
An employee working in Slack could update a CRM record directly from a conversation.
A developer could work with Salesforce capabilities through Claude.
A Salesforce user could use Agentforce Coworker within Lightning to surface insights and take action.
The underlying Salesforce platform does not disappear. Rather, the interface becomes more flexible.
This is the core idea behind AIforce.
AIforce Launches With Claudeforce, Slackforce, and Agentforce Coworker
Salesforce is launching AIforce with three major components:
1. Claudeforce
Claudeforce extends Salesforce’s partnership with Anthropic, bringing Salesforce intelligence into Claude.
Salesforce in Claude includes a prebuilt MCP server designed to provide access to Salesforce capabilities without the traditional friction associated with manually configuring integrations and authentication.
The initial release includes 37 pre-built sales skills, covering activities ranging from prospecting to pipeline hygiene. Salesforce informed their audience that Salesforce in Claude is now available to customers in beta, with analytics through Tableau. Additional capabilities for service, marketing, commerce, and industry use cases have been planned for the future.
For developers, Salesforce also highlighted a Salesforce Development plug-in for Claude Code with more than 40 skills, alongside access to Salesforce’s broader skills library.
The significance extends beyond simply connecting two products.
It demonstrates Salesforce’s broader strategy of allowing its platform to operate within the AI tools employees and developers are increasingly using.
2. Slackforce
Slackforce brings Salesforce context and intelligence into Slack.
With Slackforce Surfaces, teams can pull live information from Salesforce, Slack, and other connected tools into interactive interfaces that users can explore, filter, discuss, and act on together.
Salesforce also highlighted Slack CRM, which allows users to perform CRM actions directly from Slack.
That could include creating a Salesforce account, recording notes from a customer call, or updating records — without opening a separate CRM application.
Slackforce also extends into software development through Slack Code, allowing teams to collaborate with AI coding agents within Slack.
The broader idea is that work does not necessarily need to move between applications simply because the underlying business system is Salesforce.
3. Agentforce Coworker
The third component is Agentforce Coworker, an AI teammate built directly into Salesforce Lightning.
Agentforce Coworker can reason across account information, activity and history, surface insights and take actions within the Salesforce environment. Importantly, Salesforce says it operates within existing permissions and business rules.
In one of his keynotes, Salesforce CEO Marc Benioff stated that Agentforce Coworker was activated by 100,000 users within its first 35 days.
He also highlighted the success story of Fulton Bank, which went from zero to more than 20 Coworker use cases in production in a matter of weeks, supporting approximately 3,000 users.
This provides an interesting contrast with the external-interface approach of Claudeforce and Slackforce.
With Agentforce Coworker, Salesforce is bringing AI deeper into the existing CRM environment.
With AIforce more broadly, Salesforce is also taking CRM capabilities outside that environment.
From A Fixed UI To A Living Interface
One of the most interesting aspects of AIforce is the move away from a fixed user interface.
Traditional enterprise applications provide users with predetermined screens, fields, menus and dashboards.
AIforce introduces a more dynamic model.
Salesforce experts say users can describe the interface, view information they need, and create a customised experience around that requirement.
Instead of navigating through a fixed sequence of screens, users can increasingly ask for the information or action they need and have an AI-powered interface assemble the relevant context.
This could reduce the friction associated with navigating complex enterprise systems.
More importantly, it could change how organisations think about CRM adoption.
The question may no longer be: How do we get more employees to use Salesforce?
It may increasingly become: How do we make Salesforce available wherever our employees already work?
The Headless Toolkit: What Makes AIforce Possible?
Behind AIforce is Salesforce’s Headless Toolkit.
Salesforce describes it as an open architecture that exposes elements of the Salesforce platform through technologies including MCPs, APIs, plug-ins, skills, and developer tools.
This is an important technical development because it allows builders and developers to create customised AI experiences on top of Salesforce’s existing architecture.
It also supports the growing AgentExchange ecosystem, where partners can build and distribute interfaces, agents, applications, integrations, workflows, and actions.
The result is a more open Salesforce architecture in which the CRM does not have to be experienced exclusively through the traditional Salesforce UI.
For organisations with complex technology environments, this could create more opportunities to connect Salesforce with existing AI and productivity platforms rather than forcing users into a single interface.
AIforce Is Still Built Around Salesforce’s Core Architecture
It is important not to interpret AIforce as Salesforce becoming less important.
In many ways, the opposite is true.
AIforce is built around the information and business logic already contained within Salesforce.
Salesforce’s Agentic Enterprise architecture combines several layers:
- Data 360 provides unified and harmonised data, metadata, and memory.
- Customer 360 provides business logic, processes, permissions, and actions across functions such as sales, service, marketing, and commerce.
- Agentforce provides the digital workforce, including ready-to-deploy agents and tools for building custom agents.
- AIforce brings those capabilities into the AI interfaces where people and agents work.
This means Salesforce remains the underlying business system, while the interface becomes increasingly flexible.
The Future of Agentforce
Agentforce remains central to Salesforce’s vision.
But Dreamforce 2026 suggests that the conversation around Agentforce is evolving.
The question is no longer simply: “What can an Agentforce agent do?”
It is increasingly: “Where should people and agents interact with Salesforce, and how should Salesforce intelligence move across the organisation?”
Agentforce agents can work with Salesforce data, business rules, and workflows while AIforce provides the interface layer that allows those capabilities to reach users and agents across different environments.
That distinction is important.
An organisation might have an Agentforce agent designed to help sales teams with lead qualification.
Under a traditional model, employees might interact with that capability inside Salesforce.
Under the AIforce model, relevant Salesforce intelligence and actions could potentially surface through the tools employees already use.
The result is a more connected model of human and AI collaboration.
Security and Governance Remain Part of Salesforce’s AI Architecture
Salesforce’s positioning around AIforce places these controls directly within the architecture.
Salesforce says AIforce requests run through existing permissions and business rules, meaning agents see only the information the requesting user is permitted to see, while actions are routed back through Salesforce. The company also states that AIforce operates with Zero Data Retention, with business data used to answer a request rather than retained by the model provider.
This is particularly relevant for Australian organisations operating in regulated or data-sensitive environments.
As AI becomes capable of reading records, reasoning across information, and taking actions, businesses need to understand not just what an AI system can do, but what it is permitted to access and what it is permitted to change.
AIforce’s approach is therefore significant because Salesforce is attempting to carry its existing governance model into these new AI interfaces.
Additional Dreamforce 2026 Highlights
AIforce was the headline announcement, but it was only one part of a much broader Salesforce strategy.
Dreamforce 2026 also brought announcements around:
- Koa, Salesforce’s CRM reasoning model built on NVIDIA Nemotron
- Headless 360 and the Headless Experience Layer
- Builder Central and AI-assisted application development
- Quick Pages
- AgentExchange
- New security and privacy capabilities
- Expanded collaboration with AWS, Google Cloud, Anthropic, and NVIDIA
- Further Agentforce customer deployments
Together, these announcements point towards a common direction.
Salesforce is attempting to make its platform more agentic, composable, connected, and accessible through multiple interfaces.
That is a much broader proposition than simply adding an AI assistant to their world-class CRM.
What Does AIforce Mean for Australian Businesses?
For Australian Salesforce customers, AIforce raises some practical questions.
1. Is your Salesforce data ready for AI agents?
AI agents are only as useful as the context they can access.
Organisations should assess data quality, duplication, disconnected systems, outdated records, and knowledge sources before expanding agentic workflows.
2. Which CRM processes should become agentic?
Not every Salesforce process needs an AI agent.
The strongest candidates are likely to be processes where employees spend significant time searching for information, performing repetitive actions, coordinating across systems, or responding to predictable requests.
3. Where should employees interact with Salesforce?
The answer may no longer be “inside Salesforce”.
If teams spend most of their working day in Slack, Claude, or another productivity environment, AIforce creates an opportunity to bring Salesforce capabilities into those environments.
4. How should permissions and governance work?
As AI agents move from answering questions to taking action, organisations need clear rules around access, approvals, escalation, auditability, and human oversight.
5. What does your Salesforce architecture need to look like?
AIforce makes the quality of the underlying Salesforce architecture increasingly important.
Data models, integrations, automation, permissions, and business processes that were previously designed primarily for human users may now also need to support AI agents.
The Key Takeaway: Salesforce Comes to You
Perhaps the simplest way to understand Dreamforce 2026 is through Salesforce’s own positioning:
Salesforce comes to you.
AIforce takes the full power of Salesforce — including its data, workflows, business logic, permissions, security, and governance — and makes it available through the interfaces where people and agents already work.
Claudeforce brings Salesforce into Claude.
Slackforce brings Salesforce into Slack.
Agentforce Coworker brings an AI teammate into Salesforce Lightning.
And the Headless Toolkit gives developers and partners the architecture to build new experiences on top of the Salesforce platform.
This creates a different vision for Salesforce and a singularly unique version of a CRM.
Salesforce is no longer simply the place where employees go to update customer records.
It is increasingly becoming the business context and intelligence layer behind the work itself.
Is Your Business Ready for the Agentic Enterprise?
The announcements from Dreamforce 2026 create plenty of opportunities, but organisations should resist the temptation to implement every new capability simply because it is available.
A more practical approach is to start with the business.
Identify where employees lose time. Find the workflows that cross multiple systems. Look at where important customer or operational information is difficult to access. Assess which processes could benefit from AI-assisted reasoning or autonomous action.
Then evaluate whether the existing Salesforce architecture, data, and governance are ready to support those use cases.
That is where a structured Salesforce and Agentforce implementation strategy becomes important.
As Australia’s leading Salesforce Solution & Implementation Partner, a trusted Agentforce Enablement Partner, as well as a Registered Claude Partner, Corptec helps organisations assess their Salesforce environment, identify practical AI and automation opportunities with Agentforce, AIforce, and Claudeforce, integrate their entire CRM environment with their broader tech and AI systems, and implement solutions around measurable business requirements.
We believe that the opportunity coming out of Dreamforce 2026 is not simply to add another AI capability to Salesforce.
It is to rethink where Salesforce intelligence lives, how people access it, and how humans and AI agents work together across the enterprise.
As our Managing Director, Biret Agirtan puts it, “The question for Salesforce customers is no longer simply whether to explore Agentforce or AIforce. It is to assess whether their existing systems, data, processes, and operating models are ready to support a more connected, AI-enabled way of working.”
“Dreamforce 2026 reinforces an important point for business leaders,” remarks Nitin Ghai, Co-Founder and Principal Salesforce Consultant at Corptec. “AI transformation is not just about implementing new tools. It is about redesigning how work gets done, connecting the right data and processes, and putting governance in place so AI can deliver value in the real world. Organisations that take the time to assess their foundations and define practical use cases will be better positioned to move beyond experimentation.”
Ready to explore what AIforce, Agentforce, Claudeforce, and Salesforce’s evolving AI ecosystem could mean for your organisation?
