Salesforce used the opening keynote of Dreamforce 2026 Day 1 (September 15 to 17, Moscone Center, San Francisco) to unveil AIforce, a new interface layer for the platform.

What an attention-grabbing keynote that was. Salesforce opened Dreamforce 2026 with a clear vision for the Agentic Enterprise, showing how AI agents, humans, data, and applications are coming together to change the way businesses work.

Most Salesforce sensitive data exposure doesn't happen because someone made a dramatic mistake. It happens quietly, over time, as an org grows, new fields get added, new people get access, and nobody circles back to check whether all of that still makes sense.

In Salesforce Marketing Cloud Next, consent management occurs at the subscription level and is associated with specific contacts, such as email addresses and phone numbers, by using Communication Subscription Consent Data Model Objects (DMOs) in Data Cloud.

If you follow Salesforce even a little, you've probably heard the name AgentExchange floating around lately. Maybe you saw it mentioned at TrailblazerDX, or maybe your admin team brought it up during a planning meeting.

On April 14, 2026, HubSpot’s Spring Spotlight was big for Agent Tools, with an expanded set of Custom Workflow Actions that AI agents can now call on to get things done for users.

If you've built an AI agent package for Salesforce and sent it off for review, you already know the feeling. You wait. You refresh your dashboard. And then the email lands: "Security Review Failed."

For decades, customer service calls have started the same way. This model worked when automated phone systems had one main job: route callers to the right place. But customer expectations have changed.

Launching an AI agent or app on Salesforce AgentExchange is a major achievement, proving your solution meets Salesforce’s strict security standards. However, getting listed is only step one. 

Every integration with Salesforce Marketing Cloud starts the same way: you need an access token, and you need it fast. But Marketing Cloud API authentication isn't a checkbox; it's the front door to your customer data, and how you set it up decides whether that door is locked or left ajar.

How much of your Salesforce production data is currently visible to someone who doesn't actually need to see it? Most teams have never asked themselves that question directly, and if they did, the answer would probably surprise them.

If you're building on Salesforce and your roadmap now includes AI agents, you've probably hit the same wall a lot of ISVs are hitting right now: Agentforce actions packaging isn't quite like packaging a regular Lightning component or a chunk of Apex. 

An employee asks your internal assistant: “What’s our leave policy for people who’ve completed more than five years?” A capable model gives a fluent, confident answer. It is also wrong, because your HR policy was never in its training data.

Two years ago, Salesforce used the Dreamforce keynote stage to introduce Agentforce, the idea that AI shouldn't just assist inside your CRM; it should act. This September, that idea gets its biggest test yet.

Marketing Cloud needs customer data to power audience segmentation, personalization, journeys, and campaign activation. That data can come from Salesforce or supported external data sources connected through Data 360.

If you're running a voice AI program, you already know the dashboard trap. CSAT looks decent. Average handle time is down. Everyone in the weekly review nods and moves on. Then a customer posts on Twitter that your bot argued with them for four minutes about a refund it had no authority to issue, and nobody saw it coming because the metrics that were supposed to catch it never do.

What if your customers never had to think about upgrading your AgentExchange app again? No reminder emails. No release announcements. No waiting for admins to install the latest version. Every eligible customer simply wakes up with your newest release.

Every support team has that one metric everyone quietly fears checking on a bad week: SLA compliance. It's the number that tells you whether you're keeping the promises you made to your customers about response and resolution times.

The model is the easy part, and that’s the truth that surprises a lot of teams building voice AI. You can get an LLM sounding natural, handling interruptions gracefully, and responding in real time within a few weeks. What actually takes the work is everything around the model. The underlying infrastructure and connectivity that lets it receive a phone call, stream audio from a browser, and know who it's actually talking to.

At TrailblazerDX 2026, Salesforce co-founder Parker Harris asked a question that would have sounded absurd a few years earlier:why should anyone log into Salesforce at all?