Salesforce has unveiled AIforce, a new live interface layer designed to bring the full power of its platform to any AI interface, eliminating the need for users to open the Salesforce application directly. Announced at the company’s Dreamforce conference, AIforce aims to unlock significant value by allowing individuals and AI agents to access enterprise data, workflows, and business logic wherever they work, as detailed by Salesforce.
AIforce: A New Interface Revolution
Salesforce describes AIforce as an “interface revolution,” according to Marc Benioff, Chair and CEO of Salesforce, enabling companies to leverage their existing knowledge within Salesforce for new levels of growth and productivity. The new layer allows employees who may never have used a Salesforce dashboard to ask questions, update records, and trigger workflows directly from platforms like Claude, Slack, and Lightning. This is achieved by allowing users to compose customised live interfaces simply by describing them, providing tailored insights when and where needed, states a Salesforce announcement.
AIforce is powered by the Headless Toolkit, an open architecture exposing every element of the Salesforce platform for developers to create customised AI experiences. It also launches with specific integrations such as Claudeforce, Slackforce, and Agentforce Coworker, bringing Salesforce capabilities to popular AI-centric platforms. For instance, Elia Wallen, Founder and CEO of Engine, stated that “Slackforce makes that possible — instead of waiting on someone else’s time and expertise to turn data into a dashboard or report, anyone on our team can just describe what they need and get a real, working interface back.”
The introduction of AIforce marks a new chapter for Salesforce’s global community of “Trailblazers,” empowering them with advanced AI capabilities to transform business operations at unprecedented scale and speed.
Enhancing AI Agent Management and Security
Beyond AIforce, Salesforce also previewed two additional offerings aimed at improving the development and management of AI agents: the Enterprise AI Harness and the AI Control Plane.
The Enterprise AI Harness is a suite of technologies designed to enhance the reliability and security of customers’ AI agents, with a rollout beginning early in Salesforce’s 2028 fiscal year. It includes six feature collections, half of which focus on the business data used by AI agents. Key components include “Trusted Context” for data assets like customer records and real-time signals, as well as information quality controls like a data lineage tool, according to SiliconANGLE. Furthermore, the harness will offer “intelligent model routing” to direct agent requests to the most suitable large language model based on accuracy and cost, and provide deterministic reliability controls.
The AI Control Plane will enable administrators to centrally monitor the performance and inference costs of AI agents. It will also help ensure that agents comply with cybersecurity policies, addressing crucial governance aspects as AI integration deepens within enterprises. Rohan Kumar, Salesforce’s president and chief platform and engineering officer, highlighted that the differentiator for an enterprise will be “the trusted, proprietary context it brings to that intelligence… and its ability to securely turn that context into action.”
Industry Leaders Discuss AI’s Future and Safety
The Dreamforce keynote featured a discussion with Salesforce CEO Marc Benioff alongside the CEOs of Anthropic (Dario Amodei), NVIDIA, and Siemens (Roland Busch), focusing on how AI can reshape corporate work and industrial production, and the critical need for reliability and safety.
Dario Amodei, CEO of Anthropic, addressed the importance of AI safety, outlining a three-step plan: companies must scrutinise their own practices and invest in transparency and safety, work with the industry to raise standards, and foster international coordination. He stressed that “pacing does not mean halting model training or technical progress, but ensuring companies take adequate time to align and safeguard their models.” Amodei observed that despite rapid technological advancements, businesses are currently utilising only a small fraction of AI’s potential value, citing an example of using Claude with Salesforce data to manage sales deals effectively, as reported by R&D World.
