Salesforce has unveiled AIforce, a new live interface layer designed to integrate the full capabilities of its platform into any AI interface, announced at the Dreamforce conference. This initiative aims to empower users and AI agents to access Salesforce data, workflows, and business logic from wherever they work, without needing to open the traditional Salesforce application.
AIforce: A New Interface for Enterprise AI
AIforce is described by Salesforce as bringing all enterprise knowledge — including data, workflows, business logic, security, and governance — to any AI interface. This allows employees who may not typically use dashboards to ask questions, update records, and trigger workflows, all grounded in real Salesforce data. Agents, in turn, can read across numerous records, integrate connected systems, and take action with a complete view of business operations at speed.
Marc Benioff, Chair and CEO of Salesforce, stated that “AI is creating an interface revolution,” comparing AIforce to the shifts from DOS to graphical interfaces, the web, and mobile. He further remarked that AIforce is “the most exciting thing we’ve ever done.” The platform also empowers users to build customised live interfaces on the fly, simply by describing them within other platforms like Slack and Claude.
AIforce is launching with specific integrations, including Claudeforce, Slackforce, and Agentforce Coworker, extending Salesforce’s reach into widely used AI platforms. Elia Wallen, Founder and CEO of Engine, highlighted the impact of Slackforce for his company, noting it allows teams to get relevant customer context faster. Similarly, Brandon Bernard, Senior Vice President/Sales Technology Strategist at Fulton Bank, praised Agentforce Coworker for enabling over 20 use cases in production within weeks for approximately 3,000 users. More details on AIforce can be found on the Salesforce news website.
New Tools for Secure AI Agent Management
Beyond AIforce, Salesforce also previewed two additional offerings aimed at helping customers build and manage artificial intelligence agents more effectively. These are the Enterprise AI Harness and the AI Control Plane.
The Enterprise AI Harness is a collection of technologies designed to enhance the reliability and security of customers’ AI agents. Key features include “Trusted Context,” which combines data assets like customer records and real-time signals, and data quality controls such as a data lineage tool. It will also allow cybersecurity teams to regulate which agents can use specific datasets and how. For optimising costs, it includes an “intelligent model routing” mechanism to direct agent requests to the most suitable large language model (LLM). Rohan Kumar, Salesforce’s president and chief platform and engineering officer, underscored the importance of an enterprise’s trusted, proprietary context in differentiating its AI capabilities.
The AI Control Plane is intended to provide administrators with a central point for monitoring AI agents’ performance, managing inference costs, and ensuring adherence to cybersecurity policies, as reported by SiliconANGLE.
Industry Leaders Discuss AI’s Future and Safety
The Dreamforce keynote featured a panel discussion with Salesforce CEO Marc Benioff, Dario Amodei, CEO of Anthropic, and Roland Busch, CEO of Siemens, alongside NVIDIA. They explored how AI could reshape corporate work and industrial production, emphasising the need for reliable systems.
Marc Benioff stressed that “Models alone cannot run the enterprise,” highlighting that AI needs to be grounded in corporate data, permissions, rules, and workflows to function effectively. Dario Amodei of Anthropic addressed the critical issue of AI safety, drawing an analogy to a carmaker with a brake failure. He outlined a three-step plan for the industry: scrutinising internal practices and investing in transparency and safety, collaborating with the wider industry to raise standards, and fostering international coordination. Amodei clarified that “pacing does not mean halting model training or technical progress.” Siemens CEO Roland Busch spoke about applying AI by connecting it to engineering knowledge and the behaviour of physical products.
FAQs About Salesforce’s Dreamforce AI Announcements
-
Q: What is AIforce?
A: AIforce is a live interface layer announced by Salesforce at Dreamforce, designed to bring the full power of the Salesforce platform to any AI interface, allowing users and agents to access data and workflows without needing to open the main app.
-
Q: What other AI management tools were previewed?
A: Salesforce previewed the Enterprise AI Harness and the AI Control Plane. The Harness improves the reliability and security of AI agents, while the Control Plane allows administrators to monitor agent performance, costs, and security compliance.
-
Q: Who discussed AI safety at Dreamforce?
A: Anthropic CEO Dario Amodei joined Salesforce CEO Marc Benioff and others on stage to discuss AI safety, proposing a three-step plan involving internal scrutiny, industry-wide standard-raising, and international coordination.
