Skip to main content

AIforce: Salesforce Is Everywhere Now. Everything Is a Different Question.

Salesforce announces AIforce; image by TW with some help from Claude and ChatGPT
Salesforce put its platform inside Claude, Slack, Gemini Enterprise and Amazon Quick. Reaching every surface is not the same as governing the work that runs across them.

The News

At Dreamforce on 15 September 2026, Salesforce introduced AIforce, which it calls “a live interface layer that brings the full power of Salesforce to wherever people and agents work.” According to the announcement, agents can “reason and take action across allå the data, workflows, and logic inside Salesforce” from whichever surface the user happens to be in.

AIforce is an umbrella over five elements. Claudeforce, announced on 26 August, puts Salesforce inside Claude through an MCP server with 37 sales skills, plus a development plug-in for Claude Code. Slackforce brings Slackbot, Slack CRM, Slack Code and Slackforce Surfaces together, the latter generating live dashboards and working views from a plain-language request; Salesforce says Surfaces are available on paid and Free plans with Slackbot enabled, with live data from October. Agentforce Coworker sits inside Lightning, and Salesforce reports 100,000 users activated it within 35 days. The Headless Toolkit exposes the platform through MCP servers, APIs, plug-ins and skills, building on the Headless 360 expansion of 19 August. AgentExchange partners include Anthropic, AWS, Google and Microsoft.

Two partner announcements widened the reach. Salesforce capabilities are coming to Gemini Enterprise through a federated connector, in private preview with general availability planned for late October, and Gemini models now run in the Agentforce Reasoning Engine (Google Cloud). Salesforce context also appears in Amazon Quick. Separately, Salesforce reported “7 billion AWUs across Agentforce and Slack.”

Marc Benioff framed it as “an interface revolution” built with “Zero Data Retention,” defined as business data that “is not retained by the model provider.” In the keynote he added that “there was never any head on the apps.” Co-founder Parker Harris is quoted with “You may never log in to Salesforce again.” No prices were published, and the release states that “pricing and packaging are subject to change.”

The Bigger Picture

Two long running trends are combined in this announcement. One is behavioral: people state what they want in a conversation rather than navigating screens. The other is architectural: applications are more and more being broken into capabilities that agents can discover and call using MCP.

On the second, the field is crowded, and Salesforce is not an early bird. Plus, modularization is not exactly new (CORBA, anyone?).

  • ServiceNow made its Action Fabric MCP Server generally available on 5 May 2026, included in every Now Assist and AI Native SKU, with headless actions consuming existing Assist currency (ServiceNow).
  • SAP announced on 20 May 2026 that Claude is coming to its Business AI Platform and Joule (SAP).
  • Microsoft added MCP to Dynamics 365 Sales on 21 July 2026, centered on Copilot Studio and its own agents (Microsoft).
  • HubSpot runs connectors for Claude, ChatGPT, Gemini and Copilot (HubSpot),
  • and Zoho positions Zoho MCP as “completely model-agnostic” (Zoho).

What distinguishes AIforce, however, is its breadth and the generated interface. Four external surfaces are named, and in Slack the view itself is produced on demand rather than configured in advance. Where predefined screens lose their value, the data and data model including meta data, the permissions and the action layer gain it.

There is also an absence. AIforce names neither Microsoft 365 Copilot nor Teams, and ChatGPT appears on the Headless 360 page but not in this announcement. Microsoft shows up as a partner, not as a surface.

My Analysis and Point of View

AIforce is more a consolidation than a launch. Claudeforce, Headless 360, Slack Code and Coworker were all announced in the four weeks before Dreamforce. The new pieces are Slackforce Surfaces, Slack CRM and the Gemini Enterprise and Amazon Quick connections. Salesforce has given the whole set a name, which helps buyers see it. Still, buyers should resist counting the same capability three times because it arrived in three press releases and with different names.

Strategically the move is consistent. Salesforce accepts that it will not own every screen and concentrates on what it can control: the point where an action executes and permissions are enforced, Slack as the one surface it owns, and the business context that makes any assistant useful for work. That is a redefinition, not a retreat.

AIforce also does not undermine the seat. A person reaching Salesforce from Slack, Claude or Gemini Enterprise is still a user and still needs a license, and a license usable from every surface is easier to justify. The pressure on the price list comes from somewhere else: the job-ready agents launched the same week, which take on work that used to require a licensed person. Each task they absorb is a task no seat is needed for, and the Agentic Work Unit is not yet an answer. AWUs count “a prompt processed, a reasoning chain completed, or … a tool invoked” as the same unit, whether they are equivalent or not.

The even bigger challenge is governance. Salesforce governs what executes in Salesforce. That is real, and it is narrow. Still, Salesforce alludes to playing the ServiceNow game. Zero Data Retention binds the model provider, not the surface that displays the data, and it says nothing about what Claude, Gemini Enterprise, Amazon Quick or a Slack channel retain or log.

Surfaces are the smaller part of the problem anyway. The processes that matter most cross domains. Quote-to-cash runs through CRM, CPQ and ERP. Case resolution touches service, field service, IT and billing. When agents run those processes, several systems of record are involved, each with its own rules, identities and audit trail, and increasingly with its own agents. Salesforce cannot govern SAP's financial rules or ServiceNow's IT workflows, and it does not claim to. What is missing, however, is a coordinating layer above the domain platforms: a shared agent and identity registry, policy on which data may cross which boundary, handoff contracts at named integration points, and one audit trail that can reconstruct an outcome end to end. I made the same argument from the other direction when Zendesk declined to become a cross-system orchestrator, and the principle holds for every Titan: you cannot govern what you cannot understand.

Then there is the meter. More surfaces mean many meters on a single task: an assistant seat here, Flex Credits there, a flat Slack plan in between. No single invoice shows the whole task, and a live Surface that refreshes as data changes does not come for free either. This is the pattern I described in The Inference Paradox. Buyers do not purchase tokens, they purchase workflows, and workflows consume tokens faster than tokens get cheaper. The problem here is not a large bill. It is a bill nobody can attribute to a workflow or an outcome.

There are three things to do if AIforce is on your agenda:

  • Start from your surface, not from Salesforce's list. If your people live in Microsoft 365 or Teams, AIforce does not reach them yet. Ask for a committed roadmap before it enters a business case.
  • Get the meter in writing, surface by surface. ServiceNow has stated what is included and which currency a headless action consumes. Ask Salesforce for the same, then price one workflow end to end across all involved vendors.
  • Name where Salesforce's governance ends. Treat it as one member of a federation, define the integration points where it hands off, and decide deliberately who holds the cross-domain role. Log retention and audit parity are per-surface questions, not one question.

For Slack-heavy Salesforce customers and for organizations already standardized on Claude, Gemini Enterprise or Amazon Quick, AIforce is worth having a deeper look. Microsoft-centric, service-led and regulated organizations can wait without losing much.

The full aheadCRM research note, AIforce: Everywhere, but Not Everything, works through all of this in detail: what is new versus repackaged, the evidence behind the headline numbers, federated governance across domains, the token economics, the impact on Slackforce and Claudeforce, and the competitive read across Microsoft, ServiceNow, SAP, Oracle, Zoho, HubSpot and the model vendors. Download your free copy today.


Labels:

Comments

Post a Comment

Last Year's Top 5 Popular Posts

You are only as good as your customer remembers

As you know, I am very interested in how organizations are using business applications, which problems they do address, and how they review their success. In a next instance of these customer interviews, I had the opportunity to talk with Melissa Gordon , Executive Vice President, Enterprise Solutions at Tidal Basin about their journey with Zoho. You can watch the full interview on YouTube. Tidal Basin is a government contractor that provides various services throughout the government space, including disaster response, technology and financial services, and contact centers. Tidal Basin started with Zoho CRM and was searching for a project management tool in 2019. This was prompted by mainly two drivers. First, employees were asking for tools to help them running their projects. Second, with a focus on organizational growth and bigger projects that involved more people, Tidal Basin wanted to reduce its risk exposure and increase the efficiency of project delivery. This way, the compa...
Post a Comment
Read more

SAP Draws a Perimeter around Agentic AI and What That Means for the Rest of US

The most consequential enterprise AI governance document published this year arrived in late April with surprisingly little fanfare. SAP's updated API Policy, version 4/2026 , is a short document in plain English. The clause that is most interesting is Section 2.2.2. It restricts how autonomous and generative AI systems are permitted to interact with SAP APIs. Read literally, it has the potential to change the architecture of agentic AI projects across every SAP customer landscape. Read carefully, it is also more interesting than the lock-in headlines suggest. The policy targets a specific category of AI behavior, not AI as such. It connects to commercial mechanics that go well beyond API stability. And the literal text, in its current form, will probably not survive the next two policy revisions intact. There is a lot to unpack. I will walk through what the policy actually says, how the SAP-watching community is reading it, what the rest of the major enterprise vendors are doin...
Post a Comment
Read more

Beyond GDPR: Is MyTerms the New Standard for Enforceable Personal Data Agreements?

The news IEEE just released standard 7012-2025 for machine readable personal privacy terms , nicknamed MyTerms. MyTerms covers interactions and agreements between individuals and service providers they interact with on a network. It defines a way for personal privacy requirements to be expressed as standard-form contractual agreements.  MyTerms is intended to replace today’s “notice and consent” pattern with a standardized, machine-readable contract handshake between an individual and a service provider. The standard considers individuals true first parties who can proffer privacy terms as contractual terms, typically through an automated agent acting on their behalf. The system relies on a neutral, non-business entity that hosts a bounded set of standard-form privacy agreements. These agreements are designed to be understandable and usable in practice by humans and by machines. They must be available in plain-language human-readable form, maintain legally meaningful wording, and a...
Post a Comment
Read more

The Illusion of Value: Why Salesforce’s Agentic Work Unit is the New "Bad Query" of the AI Era

The News On February. 25, 2026, Salesforce announced a pricing and metrics update . During the company’s Q4 FY2026 earnings call, CEO Marc Benio ff, together with CMO Patrick Stokes , unveiled the Agentic Work Unit (AWU). Positioned as a metric to quantify the labor performed by autonomous digital systems, Salesforce defines an AWU as one discrete task accomplished by an AI agent. According to Salesforce, this discrete task represents the exact moment " raw intelligence is converted into real work ". It is not a fixed unit but measured as a processed prompt, a completed reasoning chain, or an invoked tool. Salesforce explicitly designed the AWU to move the industry conversation away from the raw consumption of Large Language Model (LLM) tokens. As Benioff noted, tokens only measure "how much an AI talks," whereas the AWU is intended to measure actual business execution. The scale of this rollout is massive. Salesforce reported that its platform has already processe...
Post a Comment
Read more

LLM Showdown: Comparing ChatGPT, Gemini, and Grok for Automated News Research

The analyst’s day is full of research. Now, this is the age of AI and AI is here to help, isn’t it? As everyone is talking about copilots and AI agents, why not using the tools at hand to do a little research on research. NB., no one really has a good definition of an AI agent, so this might become an additional topic for research. But I digress. Imagine the following project at hand, which is not only interesting for analysts, btw, but also for a variety of roles in the corporate world. Let’s call it vendor (competitor) monitoring. The job is the following: Research reputable sites for news about a number of vendors, relating to a set of keywords. Reputable sites are high quality news sites, high quality tech publications, high quality analyst sites and, of course the news pages of the vendors in question. Limit the time frame of the search matching to the cadence of my information requirement, e.g., “yesterday” for a daily update or “last week” for a weekly update. Provide a summary ...
Post a Comment
Read more