Dreamforce 2026 arrives as enterprise AI moves beyond experiments. The annual Salesforce event runs September 15-17 in San Francisco. A free Salesforce+ virtual program continues until September 18.
This year’s theme is the “Agentic Enterprise.” That means connecting people, data, and AI agents across daily work. For telecom and UC teams, the event deserves close attention.
Salesforce plans more than 1,600 breakout sessions and 50 keynotes. It also promises 150 hands-on trainings and 240 community roundtables. Marc Benioff will lead the opening keynote with special guests.
The biggest story is Agentforce, Salesforce’s AI agent platform. These agents can handle routine tasks, update records, and trigger workflows. They aim to reduce manual work across sales, service, IT, and operations.
However, buyers need more than polished demonstrations. They need proof from live deployments. Salesforce says Agentforce annual recurring revenue has passed $1 billion. That figure signals momentum, but value depends on execution.
Customer examples will matter. PenFed has discussed a 76-agent deployment across several business functions. It expects nearly $1.6 million in savings this year. UCLA Health also launched an Agentforce use case after eight months of preparation.
For collaboration leaders, Slack may become a central discussion point. Salesforce wants to move CRM actions into everyday conversations. Teams could review accounts, find data, and approve next steps without switching tools.
This could help sales and service teams move faster. It could also improve context during customer conversations. Yet it raises important questions about permissions, audit trails, and data accuracy.
Salesforce is also using Model Context Protocol servers, known as MCP. In simple terms, MCP helps AI tools connect to business systems. That may let Slackbot access CRM data, Tableau, Data 360, and other applications.
The approach sounds powerful for unified communications environments. Customer context could follow employees across channels and workflows. Still, enterprises must ask how this works with Microsoft Teams, Zoom, and contact centers.
Data quality will remain a central concern. AI agents can only act safely with reliable information. Companies must define ownership, access rules, retention policies, and approval paths before deployment.
Cost control also needs scrutiny. AI activity can scale quickly across large organizations. Leaders should ask how Salesforce measures agent work, errors, escalation, and business outcomes.
Dreamforce 2026 will not just showcase new products. It will test whether agentic AI can support governed enterprise work. For IT, telecom, and collaboration teams, the event offers a timely checkpoint.
The strongest sessions will show real outcomes, not just future plans. Attendees should leave with evidence on integration effort, adoption, security, and measurable value. That evidence will shape the next wave of AI-enabled workplace communications.
