Dive Brief:
- Salesforce unveiled a connective layer for its agentic enterprise architecture Tuesday at Dreamforce 2026. AIforce, an interface layer, brings the company’s Data 360, Customer 360 and Agentforce foundation to “any AI interface,” according to a press release. AIforce also lets enterprises build custom user interfaces to surface desired insights in their productivity platform of choice.
- The CRM company is combining model intelligence with Salesforce data context to create a system that works with core enterprise platforms, CEO and Chair Marc Benioff said in the press release.
- “These four layers — data, apps and semantics, agents, and interface — that is what we want to transform right now,” Benioff said Tuesday during the company's Dreamforce keynote.
Dive Insight:
Salesforce’s new interface layer builds on its agentic investments as the company continues to pursue enterprise spend.
Global AI spend is projected to hit $2.59 trillion this year as enterprises double their spending on generative and agentic AI, according to a Gartner report. Salesforce has leaned on its agentic platform Agentforce to attract enterprise dollars and maintain revenue growth.
Benioff teased the launch of AIforce, describing it as an “enterprise harness,” during the company’s Q2 2027 earnings call in August.
AIforce sits on top of Salesforce’s core systems, including Agentforce, and connects to companies’ AI productivity tools while maintaining the context and permissions that previously existed within the Salesforce environment, Benioff said during the earnings call. While employees and agents can surface answers from Salesforce data, the system is built with a zero data retention policy, he added.
“People don’t have to come to Salesforce to get work done,” Benioff said during the earnings call. “Salesforce comes to them, and Claude and Slack and ChatGPT and Teams, wherever they want to work. And it’s a shift from software as the interface to software powering every interface.”
The new interface layer could help reduce the time customers spend waiting on Salesforce administrators to fix issues and help them get to value faster, Terra Higginson, distinguished analyst at Info-Tech Research Group, said in an email.
“This takes Salesforce’s idea of headless CRM to the next level and moves it closer to what I would think of as ambient CRM,” Higginson said. “People can use CRM wherever it makes the most sense to them, while the data, logic, permissions and workflows that run the business stay within Salesforce.”
However, AIforce likely won’t come cheap as AI potentially introduces a new layer of cost and it doesn’t remove traditional data challenges, she said. Customers will still need clean, connected and governed data underneath.
Additionally, while AIforce promises that users can build their own interfaces, UI design requires additional skills that may present challenges to customers, Igor Ikonnikov, advisory fellow at Info-Tech Research Group, said in an email. While enabling UI build for a technically-savvy business user is a good idea in theory, it may be hard to expect less technically-proficient users to understand the whole system architecture, he added.
Salesforce also announced expanded partnerships with Google Cloud and AWS at Dreamforce. Enterprises can surface business context in tools such as Amazon Quick, while AWS frontier agents will be added to Slack, as a result of the updated partnership.
Meanwhile, Hyperforce — Salesforce’ next-generation infrastructure for public cloud — will run on Google Cloud and enterprises will be able to surface Salesforce data directly within Gemini Enterprise.
