News Analysis
On October 14, at San Francisco's Moscone Center, Salesforce CEO Marc Benioff stood before 50,000 of the Salesforce faithful, ready to preach the gospel of Agentic AI.
This was Dreamforce, Salesforce's user conference, last week. Here, Benioff urged every enterprise leader to reimagine a world where AI elevates human potential like never before. "We're entering a new era where every employee can have an AI agent working alongside them — co-piloting decisions, automating tasks and amplifying human creativity." While Benioff insisted from one side of his mouth that AI wasn't going to take jobs, he spoke about "leaner, more efficient organizations" out the other.
What more can we expect from a CEO who believes that everyone has a home, an office and Airstream interconnected in their yard and has laid off 18,000 workers in the past 3 years?
Benioff then went on to introduce Agentforce 360, built upon Data 360 (the renamed Data Cloud), the next evolution of the Salesforce platform. It connects humans, agents and data in one "trusted" platform. The data provides context, Benioff argued. "If you don't have your data, you don't have your context."
Salesforce also announced deeper partnerships to support this ecosystem, notably an expanded collaboration with Anthropic, making Claude a preferred AI for regulated industries, and a new initiative with OpenAI for deeper integrations into ChatGPT.
Why Enterprises Should Care
Benioff wasn't just pitching product features. He came with numbers. Companies integrating Salesforce data into Slack have cut their deal stage times by up to 60%. Some have seen productivity gains exceeding that. These aren't marginal improvements. If they're real, they represent the kind of efficiency leaps that justify moving budget around and getting C-suite buy-in.
Most enterprise leaders are drowning in application sprawl and fragmented workflows. Employees toggle between a dozen browser tabs just to get simple tasks done. Benioff's pitch is that Slack conversations, customer data and AI agents can work together in a single workspace. He framed it as eliminating toil. Michael Dell, CEO of Dell, joined the keynote to reveal they cut their supplier onboarding process from 60 days down to 20 days. That's a 66% reduction.
The promise is one we've heard before: AI does the drudge work faster so workers can spend time on higher-value stuff.
"Since agentic AI is blurring the lines that separate applications, data and individual tasks, it's a logical next step to shift our mindset toward optimizing our entire operational processes with agentic flows and technologies," Constellation Research's VP and senior analyst Martin Schneider told Reworked.
What Slack Actually Wants to Be
Benioff proposed that Slack isn't a messaging app anymore. He called Slack "your agentic OS" — a variation on its recent drive to be the work operating system — which brings together everything an enterprise needs into one unified place to drive work at the speed of AI. This shifts how work gets done, transforming Slack into a single, conversational workspace where people, AI and agents collaborate, with the full context of conversation and data, to move work forward faster, together.
Slack aims to be the interface layer between employees and everything else — CRM data, HR systems, engineering tools, AI agents — the whole stack. The underlying Data 360 positions it as the intelligence layer providing agents with trusted, contextual data.
Agentforce 360 provides the Agentforce Builder tool and Agent Script language to help customers design autonomous workflows with built-in governance and oversight capabilities. These weren't hypotheticals. Benioff used Anthropic as an example, which accelerated deal cycles by 60% and saved $4.5 million in deal operations using Slack. He positioned this as early evidence of a broader shift, "One of the single biggest unlocks in human potential that we'll probably ever see."
He described Salesforce at the center of the "Agentic AI Revolution" and framed it as a shift from reactive systems to "intelligent, self-learning networks" where "humans and agents drive customer success together."
Slackbot Gets Serious
In another keynote, Slack CEO Denise Dresser built on Benioff's vision with a more practical view of daily productivity. She opened by emphasizing what employees actually want from technology: "We want to have a conversation with technology. Frankly, we want to ask anything and find everything immediately at our fingertips," she said. That conversational interface is Slackbot, an AI companion grounded in company data.
Dresser highlighted practical use cases. Over the past year, Slack shipped over 1,000 new features — writing assistants, real-time translations, thread summaries, AI huddles that remind teams to follow up on commitments. Each one is designed to return time to employees. "We've had the gift of time coming back to you because we've done thread summaries and agent summaries and channel summaries," Dresser said.
Then Dresser talked about Slack Enterprise Search, launched in March 2025, which "brings us that beautiful conversational interface we've come to love right into the flow of work, giving access to the entirety of the knowledge of my company." Not just Salesforce but also files from Google Drive, messages from Teams, data from Asana and GitHub — all accessible without leaving Slack. "No more context-switching. No more hunting through multiple systems," she said.
The new version of Slackbot can handle multi-step tasks like drafting emails, creating prep documents, coordinating team availability and updating CRM opportunities. Dresser described the updated Slack as "grounded in your conversations, your files, your workflows, your apps, and it's right there in the flow of work."
Two Views of AI Transformation
Benioff and Dresser approached the transformation from different angles. Benioff focused on business outcomes and enterprise-scale impact — the 60% reduction in deal cycles, the $4.5 million savings, the shift from fragmented tools to unified workflows. Dresser zoomed in on employee experience — the daily friction eliminated, the time reclaimed, the conversational simplicity of asking a question and getting an answer.
The combined vision: the Agentforce 360 platform, powered by Data 360, provides the intelligence, while Slack provides the operational interface. There were no contradictions. Just complementary views of the same future. Benioff showed what's possible at scale. Dresser showed how it works in practice.
What Enterprises Need to Consider
The Dreamforce keynotes painted a picture of Slack as an "agentic operating system,” a phrase that sounds like marketing but was presented as a real architectural shift. Instead of employees navigating between systems, they work inside a conversational interface where AI agents, company data (supported by Data 360), and human teammates converge.
Schneider put it simply: "Agentic AI is the nexus between humans and agents."
Salesforce is betting on Agentforce 360 to bridge the gap between AI prototypes and production-ready enterprise systems. Companies like Anthropic are apparently already adopting this and seeing measurable results.
Conference
Oct
20
Gartner IT Symposium/Xpo Orlando 2025
Register
Conference
Oct
27
Gartner HR Symposium/Xpo Orlando 2025
Register
The broader question Benioff posed remains: "How do I transform my company? How do I transform the products and services we offer to meet this moment?"
The answer, according to Salesforce, starts with rethinking the interface layer between employees and work itself. And right now, Slack is positioning itself as that layer.
For enterprise leaders, this matters because it's not about buying another tool. It's about reconsidering how work flows through the organization — and whether a unified conversational layer can actually deliver the efficiency gains Salesforce claims. The numbers look compelling. Whether they translate across different industries and company sizes remains to be seen.
Editor's Note: Catch up on more news from the collaboration and productivity space:
- The Future of Team Collaboration According to Coda, Microsoft Loop, Notion and Slite — A look at how four generative AI-powered collaboration platforms aim to transform team productivity.
- ServiceNow and Salesforce Fight to Be the Center of AI Agent Operations — Salesforce and ServiceNow move into each other’s domains, but their eyes are on a much bigger prize.
- Google Opens Workspace to Claude, Part of Open Ecosystem Approach to AI — Google is unleashing third-party AIs like Anthropic’s Claude into Workspace to challenge Microsoft’s AI power grab. Could this be the future of work?
