You Might Not Log Into Salesforce Anymore
The single biggest theme from Dreamforce 2026 is bringing Salesforce to you, wherever you work. The company unveiled 'AIforce', a new layer designed to let you interact with all your CRM data and workflows from other applications, most notably Slack and the AI assistant
Claude. This isn't just about getting notifications; it's about updating records, triggering complex processes, and asking questions in plain English without ever opening a Salesforce dashboard. For a sales rep, this could mean creating a new account record directly from a Slack conversation. For a service agent, it might mean getting a customer's entire history summarized by an AI assistant without navigating through multiple screens. The long-term vision is clear: the CRM becomes an intelligent engine running in the background, rather than a destination you have to visit.
Slack Becomes Your New Command Center
The idea of leaving the main Salesforce app is powered by a supercharged Slack. A new feature called 'Slackforce Surfaces' allows you to ask the Slackbot to create live, interactive dashboards and reports directly within a channel. Imagine asking, "Show me my open high-priority cases for this quarter," and getting a dynamic chart you can filter and discuss with your team right in Slack. This 'Surface' stays connected to the source data, so it's always up-to-date, unlike a static screenshot or export. This deep integration also means that conversations about an account or service ticket can now happen in dedicated 'Salesforce Channels' that are directly tied to the CRM record, keeping everything organized and accessible. The goal is to reduce context switching and make data-driven decisions part of the natural flow of conversation.
Einstein AI Gets Way More Conversational
Salesforce's AI, Einstein, is shifting from a background analyst to a conversational partner. The centerpiece of this evolution is the 'Einstein Copilot', a conversational AI assistant designed to understand natural language. Instead of clicking through reports, you can simply ask questions. For developers and admins, the new 'Einstein Copilot Studio' allows them to build, customize, and control these AI interactions. This means companies can create specific 'skills' for the Copilot, like guiding a user through a complex quoting process or summarizing the last six months of activity on a key account. The focus is on making AI a practical tool that anyone can interact with, not just data scientists. This builds on existing capabilities like lead scoring and activity capture, making them more accessible through simple conversation.
Automation Is Triggered by More Than Just Clicks
Automation is getting smarter by tapping into more diverse data. With Data Cloud now more deeply embedded in the core platform, Salesforce Flows—the tool used to build automations—can be triggered by real-time data from various systems, not just a change in a Salesforce field. For example, a high-value customer interacting with a knowledge base article on your website could automatically trigger a Flow that creates a follow-up task for their account manager. Salesforce also announced it would provide free Data Cloud and Tableau licenses for many customers, encouraging more users to connect disparate data sources and find insights. This pushes automation beyond simple record updates and toward creating proactive, data-driven actions that can anticipate customer needs or internal workflow bottlenecks.
AI Becomes a 'Coworker' in Your Workspace
The concept of 'Agentforce' has evolved, with Salesforce introducing the 'Agentforce Coworker', described as an AI teammate that lives within your Lightning workspace. The idea is that these AI agents will collaborate with human users, supporting rather than replacing them. For instance, an AI agent could help a sales rep by preparing call notes, suggesting next steps based on recent interactions, or flagging an at-risk deal. In a service context, it might handle initial customer triage or find relevant support articles for a human agent. This 'multiplayer' approach, where humans and AI agents work together, was a consistent theme, aiming to handle repetitive tasks and surface critical information so that employees can focus on more strategic work.
