The Helpful: Your New AI Coworker
The main event at Dreamforce 2026 was the unveiling of AIforce, a new interface layer designed to let employees use Salesforce data and functions from other applications like Slack or Google Gemini. The pitch is simple and powerful: instead of logging
into a rigid system, AI agents bring the necessary information and actions to you, wherever you work. Imagine an AI that can scan hundreds of customer records, summarize the key issues, and draft a response without you ever opening a new tab. For sales, service, and marketing teams, this promises a massive productivity boost. One customer, Adecco Group, reported placing 20,000 more people this year thanks to the time savings from AI agents. The goal, as Salesforce sees it, is to create an “interface revolution” where AI makes complex software accessible and dynamic, freeing up humans to focus on relationships and strategy.
The Unsettling: Trust, Jobs, and Control
While the benefits are clear, the undercurrent of the conference is a palpable sense of anxiety. The idea of AI agents autonomously performing tasks that once defined entire jobs naturally raises questions about job security. If an AI can resolve 71% of customer cases on its own, as one company showcased, what happens to the support staff? Furthermore, as AI agents get woven deeper into business operations, the risks multiply. Concerns around governance, security, and data privacy are paramount. When an AI makes a mistake, who is accountable? How do you ensure AI agents, which can now be built by anyone, adhere to company policy and data permissions? Salesforce is addressing this with features like a “Zero Data Retention” policy and an “AI Harness” for data governance, but the sheer speed and scale of change leaves many business leaders questioning if they can keep up. The conversation has shifted from a fascination with AI demos to a sharp focus on measurable value, cost, and risk.
The Search for a Balanced Future
Salesforce is attempting to navigate this tension by positioning itself as the trusted layer for enterprise AI. The strategy is to provide the secure data and business logic that outside AI models can use safely. By partnering with companies like NVIDIA, Google, and AWS, Salesforce is trying to build an ecosystem where powerful AI models, like its new Koa reasoning model, can operate with guardrails. The company's leaders are framing the AI transition as a collaborative one, where technology happens “with people, not to people.” However, the reality is that the shift to an “Agentic Enterprise” places a huge burden on companies to redefine roles, establish rigorous governance, and manage the exponential increase in AI-driven changes. The success of this new era depends less on the power of the AI itself and more on the ability of organizations to implement it responsibly.
