AI, Agents, and Financial Services: Lessons from Dreamforce 2025

We just returned from Dreamforce 2025, and even days later, we still can’t get over the experience. Walking through the sessions, keynotes, and demonstrations, one idea kept standing out: AI is no longer in the background. It’s actively shaping how work gets done, guiding decisions, and helping teams act before challenges appear.

Dreamforce 2025 offered a front-row view of how AI is shaping the future of financial services. For those who couldn’t attend, it’s easy to imagine it as another tech conference with flashy demos and keynote speeches, but the reality was different. 

Last year was about imagination. The art of the possible. AI was something that would one day transform how we work. This year felt different. The energy at Dreamforce wasn’t about what could happen but about what is happening. There wasn’t a single person who said, “I’m not thinking about AI.”

Here are the six most impactful takeaways, along with actionable insights for financial institutions.

1. From AI Hype to Making Things Happen

The AI hype cycle is starting to balance out, and that’s a good thing. Use cases are becoming more grounded, focused on what’s possible now rather than chasing the most complex ideas first. Those who start small, learn fast, and iterate are the ones pulling ahead.

Across financial services, companies are moving from testing AI to scaling it. The conversation has shifted from “how do we start?” to “how do we go faster, safely, and with trust?” Agentforce 360 going GA this year is a clear signal: AI agents are ready for practical work, even if some capabilities are still in development. This incremental approach lets teams see value immediately and grow their adoption as the technology matures.

2. The Agentic Enterprise Becomes Real

Salesforce described the agentic enterprise as the next evolution after cloud → mobile → social → predictive. And this year, we’re starting to see what that looks like in practice.

Agents today aren’t replacing people or managing entire workflows. They’re showing up inside daily work, summarizing client notes, surfacing insights, and helping employees take the next best action.

For financial services, this might look like:

• Advisors using AI copilots to prepare pre-meeting summaries, identify client needs, or leave speaking notes post-meeting, with agents updating opportunities and follow-up reminders.
  
• Relationship managers receiving real-time insights in Slack from Data 360, showing which clients are due for outreach and providing key summaries.
  
• Service teams blending AI-generated suggestions with compliance-approved messaging to create better client experiences.
  

The takeaway: small, useful steps are already letting people and AI work together naturally. AI isn’t a standalone tool; it’s becoming a core part of how work gets done.

3. Data 360 Takes Center Stage

“If you don’t get your data right, you won’t get AI right.” That point hit especially close to home for financial institutions, where volumes of disparate and unstructured data are the norm.

Salesforce reintroduced Salesforce Data Cloud as Data 360, positioning it as the intelligence layer of the agentic enterprise. It’s where context, compliance, and connectivity come together to make AI function reliably.

For banks and wealth firms, the value comes from:

• A unified, real-time view of every customer and household
  
• Built-in governance that ensures insights are compliant and explainable
  
• Context-rich data that turns AI from guesswork into guidance
  

Best practice is to harmonize and segment in Data 360 and activate that data across multiple channels. No more point-to-point connections or redundant steps. Do it once, then use it many times, whether in Marketing Cloud for personalized engagement, Financial Services Cloud to enhance client records, or targeted call-downs and voice experiences.

The opportunity is clear: build your data once, and make every interaction smarter and more relevant in real time.

4. Salesforce and OpenAI Raise the Bar

Dreamforce’s biggest announcement was the expanded partnership between Salesforce and OpenAI, connecting Agentforce 360 with GPT-5 to bring AI directly into the workflow.

For financial institutions, this is about accessibility. AI is no longer something you log into; it comes to you. Employees can ask questions, surface insights, and act on data wherever they are.

• A relationship manager can check client updates through ChatGPT without opening a dashboard.
  
• A service agent can get summaries or recommendations directly in Slack.
  
• A wealth advisor can generate planning notes or next steps right inside Salesforce.
  

It’s the same trusted Data 360 context, just delivered more naturally and conversationally. Work is becoming conversational, and insights are available exactly when needed. For banks and wealth firms, that means faster decisions, more proactive client service, and a direct line between intelligence and action.

5. Trust and Governance Lead the Way

Financial institutions can’t afford to treat explainability as optional. Regulators are asking hard questions: What model are you using? How is it trained? How is data masked and audited? Who reviews outputs before decisions are made?

Salesforce addressed this by embedding governance, observability, and transparency into Agentforce 360 and Data 360. The observability layer allows firms to trace AI actions end-to-end, view which model was used, and apply controls before agents act.

Model flexibility also means organizations can bring their own LLMs while maintaining visibility, security, and compliance. With these enhancements, Data 360 allows Agentforce Voice to extend AI beyond text-based inputs to support natural, contextual conversations delivered in a human-like way.

For banks and wealth firms, this is critical. Compliance and risk teams get the visibility they need to approve AI use cases confidently. Innovation moves forward without sacrificing safety, and trust is no longer just a promise. It’s a system embedded in the platform.

6. Humans and Agents, Better Together

The theme throughout Dreamforce was partnership, not replacement. People and AI are working together to deliver better outcomes faster, while keeping guardrails in place for unexpected turns.

As actor Matthew McConaughey noted in a recent Agentforce video: AI doesn’t feel. AI doesn’t care. But people do. That is what sets us apart.

The goal isn’t agents that think like humans, but agents that enable human growth. AI handles repetitive tasks, freeing teams to think deeper, act faster, and connect more meaningfully. Together, humans and agents move faster, serve smarter, and strengthen trust.

AI in Action, People in Control

For banking and wealth management, success isn’t about chasing the next shiny AI tool. It’s about building on strong data foundations, ensuring governance, and enabling people to shine.

Some capabilities of Agentforce 360 are still in development, but even today, the platform can transform workflows and client interactions. Firms that start now, with small, meaningful use cases, will lead the way in the agentic enterprise.

Dreamforce 2025 wasn’t about distant possibilities. It was a clear signal: the future of finance is here, and it’s agentic. Those who align data, trust, and people now will be the ones shaping what comes next.