Salesforce launches AIforce, fleshes out platform play

Salesforce launched AIforce, which operates as an interface layer that brings the company’s platform to an AI interface.

The effort, which was outlined at Dreamforce, is the latest chapter in Salesforce’s plan to be a platform that expands beyond CRM to multiple enterprise workflows.

• Salesforce launches Koa CRM reasoning model built on Nvidia's Nemotron
• Salesforce still has pricing work to do
• Salesforce aims to be your enterprise AI harness, touts AIforce
• Salesforce Q2 shines amid AI demand

Salesforce CEO Marc Benioff had teased AIforce during the company’s second quarter earnings call. The launch of AIforce is the follow-up to the idea that you’ll never have to login to Salesforce directly to use its data, workflows, business logic and governance to use the platform. “We are combining model intelligence with all the context that customers have built into Salesforce to create an intelligent, dynamic, composable system,” said Benioff.

During his keynote, Benioff said the idea behind using AIforce to bring Salesforce to other interfaces is about "unleashing trapped value" via metadata and previously siloed systems that can be linked.

Key points about AIforce:

• The layer enables users to use Salesforce’s platform in Anthropic’s Claude, Slack and other surfaces to update records, ask questions and trigger workflows.
• Agents will be able to work across multiple connected systems using AIforce.
• AIforce features Zero Data Retention to leverage frontier models without retaining enterprise data and context.
• The architecture behind AIforce includes Data 360, Customer 360 and Agentforce.

Salesforce as enterprise AI enabler

Benioff's keynote revolved around positioning Salesforce as the center of enterprise AI and the glue that leads to transformation.

He said:

"We have to do more to help you accelerate that transformation. But one of the key parts about that is this: models alone cannot run the enterprise. Models alone are not going to show us what's going to what's possible. Models also are very probabilistic. They kind of know what's going on, but they're not grounded in a single set of truth like your traditional Salesforce apps."

Salesforce's plan is to bridge deterministic data and probabilistic models. "How can you bring that probabilistic world, the AI world, the deterministic world? The deterministic world is where all that corporate data is, where the single source of truth is. That is the question that has been on our mind, and that is the bridge that we want to cross," said Benioff. "These four layers: data, apps, and semantics; agents and interface-that is what we want to transform right now."

Benioff said AI has really brought an interface revolution to the enterprise. He added that the interface is alive.

AIforce is that layer that can bridge AI front ends and governed data on the back end.

To align its pricing with AIforce, Salesforce launched another Max plan that includes a full buffet of AI and layers to deploy.

The Dreamforce news barrage

Salesforce has had a steady drumbeat of news leading up to Dreamforce. The most notable were Claudeforce, which brings Salesforce to Anthropic's Claude.

Here's a look at what Salesforce is announcing:

An expanded Google Cloud partnership. Salesforce said it has expanded its partnership with Google Cloud so AI agents on both platforms can work together. Key parts of the deal include:

• Hyperforce will run on Google Cloud. As a result, enterprises will be able to leverage Google Cloud infrastructure and data as well as Gemini Enterprise.
• Enterprise can surface Salesforce directly within Gemini Enterprise without integration via Salesforce’s headless architecture.
• Tableau analytics will be available in Gemini Enterprise.
• Gemini models will be available in Agentforce.
• Commerce Cloud will be integrated with Universal Commerce Protocol and Google Checkout.

Salesforce will be available in Amazon Quick with more AWS integrations. Salesforce announced several new integrations with AWS services. Salesforce data, context and actions will be available in Amazon Quick and AWS agents will be available in Slack. There will also be support for bidirectional voice between Agentforce Voice and Amazon Connect Customer.

In addition, Agentforce customers will get access to Amazon Bedrock’s full roster of models.

Siemens Agentforce integration. Salesforce and Siemens said Siemens Teamcenter will work with Agentforce. The combination will meld Agentforce with Siemens Teamcenter Service Lifecycle Management, which is used for industrial sales, service and customer workflows.

Prepackaged AI agents. Salesforce launched a series of "job-ready" AI agents that can take on work across sales, service, commerce, employee experience and back office functions.

The agents, powered by Agentforce, are designed to give enterprises a running start. Salesforce also gave all of these agents names. Here's the lineup:

• Casey, a help agent is focused on omnichannel customer service issues and is generally available.
• Paige, an IT and HR service agent is generally available.
• Carter, a shopping agent for product discovery and comparison, is generally available.
• Hunter, an outbound sales agent is in pilot now and generally available in November.
• Marshall, a supply chain agent is generally available.
• Piper, an inbound pip gen agent for B2B sales and marketing teams, is generally available.
• Fin, a customer agent designed for complex issues, is generally available. Salesforce recently completed its acquisition of Fin. See: Salesforce Acquires Fin and Leaves Us Reading Between the Agentic Lines

Enterprises can customize the agents and rename them. According to Salesforce, these agents are driving outcomes by resolving inquiries, building pipeline and resolving issues autonomously.

Enterprise AI Harness. Salesforce announced its Enterprise AI Harness, which is designed to give AI agents the ability to reason, plan, take actions and operate with business context. The harness enables enterprises to keep one set of context for multiple AI agents.

Salesforce's Enterprise AI Harness includes tools for context, agency, action, governance, security and models. The architecture can put together customer relationships, processes and controls. The harness leverages most of Salesforce's portfolio including Data 360, Informatica, MuleSoft and Agent Fabric, Agentforce, Salesforce Guardian and Salesforce Platform.

The components of the Salesforce Enterprise AI Harness are available today, but the unified experience and new capabilities will land in early fiscal 2028. Pricing will be announced when generally available.

Slackforce Surfaces. In keeping with its headless vibe, Salesforce launched Slackforce Surfaces, which brings Salesforce data and applications to Slack. In a nutshell, Slackforce Surfaces brings the context and work to Slack via Slackbot natural language. Data refreshes in real-time and Slackbot can build dashboards, reports and decks via conversation.

Salesforce said Slackforce Surfaces is available for customers of multiple plans who have enabled Slackbot. Live data features roll out in October.