Dreamforce 2026 made one thing clear: the conversation about artificial intelligence has changed.

For the past few years, businesses have been asking what generative AI can do. This year, Salesforce shifted the focus to a more practical question: how can AI work safely with real business data, follow existing processes and deliver measurable results?

Across the Dreamforce program, the message was consistent. AI is moving beyond isolated chatbots and pilots. It is becoming part of the operating layer of the Salesforce ecosystem, connected to customer data, workflows, permissions and the systems employees already use.

Here are the developments we believe matter most for Australian businesses.

 

1. AIforce connects AI to the way businesses work

AIforce was one of the major themes of the Dreamforce 2026 main keynote. The concept brings humans, AI agents, data and business applications together across the places where work happens.

That distinction matters. An AI model may be able to generate a convincing answer, but without access to reliable business context, the answer may not reflect your customers, pipeline, financial position or internal controls. AIforce is intended to ground AI in the data, permissions and processes already managed within the Salesforce ecosystem.

For business leaders, the real opportunity is not simply faster content generation. It is enabling people and agents to take useful action while maintaining governance and visibility.

 

2. Salesforce in Claude brings CRM into the conversation

The arrival of Salesforce in Claude was one of the most tangible examples of this shift. The beta integration brings Salesforce accounts, opportunities and pipeline information into Claude, with pre-built skills for activities such as account research, call preparation, pipeline review and CRM updates.

Importantly, it respects the permissions already configured in Salesforce and, by default, asks the user to approve proposed changes before writing them back to the CRM.

This points to a broader change in how people may interact with Salesforce. Instead of moving between screens, reports, email and notes to prepare for a meeting or review a deal, users can work through a conversational interface that draws on authorised business information.

However, the quality of the outcome will still depend on the quality of the underlying Salesforce environment. If customer data is incomplete, processes are inconsistent or access controls are unclear, adding an AI interface will not solve the foundation problem.

 

3. Data 360 provides the context AI needs

Dreamforce also reinforced the role of Data 360 as the context layer behind useful enterprise AI. Agents need more than a prompt: they need accurate, connected and timely information.

For many organisations, this is where the biggest challenge sits. Sales may live in Salesforce, service information in another platform, finance in an accounting system and critical operational details in spreadsheets. When those systems tell different versions of the customer story, both people and AI spend time reconciling rather than acting.

Connecting that information creates a stronger foundation for automation, analytics and decision-making. It also makes the familiar promise of Customer 360 more commercially meaningful: a shared view of the customer that can extend from the first enquiry through delivery, invoicing, payment and profitability.

 

4. The focus is shifting from pilots to measurable return

The Dreamforce agenda repeatedly returned to a practical theme: moving from pilot to profit.

The businesses that gain the most from agentic AI will not necessarily be those that launch the greatest number of agents. They will be the ones that select valuable use cases, prepare their data, define the right controls and measure the outcome.

Before investing in a new agent, leaders should be able to answer:

• Which business problem are we solving?
• What data and systems will the agent need?
• Which decisions can it make, and which require human approval?
• How will we measure time saved, risk reduced, revenue gained or customer outcomes improved?
• Who will own the agent after launch?

This is especially important for finance leaders. A promising demonstration is not the same as a reliable business process. Governance, auditability and return on investment must be designed in from the beginning.

We explored this shift from CRM to execution layer earlier this year. Dreamforce 2026 confirmed the direction has only accelerated.

 

5. Trust and human oversight remain essential

Dreamforce 2026 also placed trust, security and the human role firmly within the AI conversation. As agents gain access to business data and become capable of acting, permissions and oversight become operational requirements rather than abstract principles.

The strongest approach is not to choose between people and AI. It is to design a model in which AI handles appropriate research, preparation and repetitive work, while people retain control over sensitive decisions and exceptions.

 

What Dreamforce 2026 means for your Salesforce roadmap

The technology is moving quickly, but the best next step may not be to deploy an agent immediately.

For some businesses, the priority will be connecting sales, operations and finance. For others, it will be improving data quality, simplifying a process, strengthening permissions or defining a clear AI use case with measurable value.

The important point is that AI readiness is closely connected to business-system readiness. A well-designed Salesforce environment gives AI the context and controls it needs to become useful, rather than simply impressive.

 

Continue the conversation at True Blue Blazing Melbourne

Swan Technology Solutions is proud to sponsor True Blue Blazing 2026 in Melbourne on Tuesday, 20 October at Marvel Stadium.

At 11:45 am, Jo Swadling will present Salesforce to Finance: The True Customer 360. The session will explore how organisations can connect Salesforce with finance and compare three practical approaches (partial native, connector and custom) across cost, delivery time and ongoing maintenance.

It is a natural continuation of one of Dreamforce’s most important messages. AI and automation are only as useful as the business data and processes beneath them. A true Customer 360 should not end when a deal is marked Closed Won; it should connect the full journey through operations and financial outcomes.

If you are considering how Salesforce, finance and AI should fit together in your organisation, join us at True Blue Blazing to explore the options and trade-offs in practical terms.

Event	True Blue Blazing 2026
Date	Tuesday, 20 October 2026
Location	Marvel Stadium, Melbourne
Session	Salesforce to Finance: The True Customer 360
Speaker	Jo Swadling, Swan Technology Solutions
Time	11:45 am

 

Ready to turn the announcements into a practical roadmap?

Swan Technology Solutions helps businesses make the most of Salesforce, from prospective customers to profit and loss. We connect sales, operations and finance on one platform so teams can reduce rework, improve visibility and build a stronger foundation for automation and AI.

Talk to our team about where to begin.