Salesforce Inc. (NYSE: CRM) is handing Anthropic’s Claude an unusually central role across its artificial intelligence strategy, integrating the model into Salesforce applications while simultaneously making Salesforce data and workflows accessible directly from Claude.
The expanded partnership, called Claudeforce, introduces Salesforce in Claude with 37 prebuilt sales skills capable of working with live revenue data, updating pipelines and performing governed actions. At the same time, Claude is becoming deeply embedded inside Agentforce and Slack, where Salesforce says it is now the default model for Slack AI, Slackbot, Agentforce Coworker and several internal developer tools.
The commercial timing is important. Salesforce reported Agentforce annual recurring revenue above $1.5 billion in its fiscal second quarter, up more than 240% year over year, while combined Agentforce and Data 360 annual recurring revenue approached $3.9 billion. Salesforce also said Slack generated its fastest quarterly net-new annual order value growth since the company acquired it.
Investors responded strongly to the broader earnings and AI package. Salesforce shares closed 22.58% higher at $252.05 on August 27, the first full trading session after the results and expanded Anthropic partnership, compared with $205.62 on August 26. The stock remained at $256.93 by the September 2 close, keeping most of that revaluation intact.
What exactly does Claudeforce change for Salesforce customers?
Claudeforce works in both directions.
Salesforce in Claude allows users to access Salesforce data and workflows without first navigating conventional CRM screens. The initial plugin includes 37 sales-focused skills covering functions such as meeting preparation, pipeline review and deal-health assessment. Claude can reason over business context and then route actions through Salesforce so established permissions and business rules continue to apply.
Claude is also being embedded inside Salesforce products. It is available as a reasoning model within Agentforce, powers Agentforce Coworker by default and can operate within the Salesforce Trust Boundary through Amazon Bedrock for customers with more demanding security requirements.
This moves the partnership beyond adding another chatbot to a CRM interface. Salesforce is effectively separating enterprise intelligence from the conventional user interface, allowing employees to ask Claude to reason across customer information and then use Salesforce’s workflow system to carry out permitted actions.
That architecture could matter if conversational interfaces become a primary way employees interact with enterprise software. Instead of opening several applications and manually navigating records, users may increasingly describe an objective while AI agents determine which data, workflow or application needs to be invoked.
Why is Slack becoming strategically more important inside Salesforce’s AI architecture?
Slack may be emerging as one of the most valuable pieces of the Claudeforce strategy because it provides an existing conversational environment where humans, business applications and AI agents can interact.
Salesforce says Claude will power Slackbot and serve as the intelligent backbone of several Slack experiences. Employees can potentially discuss a problem in a channel, involve an AI agent in the conversation and execute governed Salesforce actions without leaving the collaborative environment.
That is materially different from Slack’s original positioning as workplace messaging software.
Salesforce acquired Slack for almost $28 billion in 2021, a price that attracted persistent questions about whether the acquisition could ever generate adequate strategic returns. Agentic AI potentially changes the argument because Slack can become an interface layer connecting users with data and AI-driven actions across the broader Salesforce platform.
Recent operating metrics support at least part of that thesis. Salesforce said Slackbot users increased more than 150% quarter over quarter during the fiscal second quarter and that Slack recorded its fastest quarterly net-new annual order value growth since the acquisition.
Those numbers do not prove that AI will justify the original Slack acquisition price. They do indicate that Slack is becoming more closely tied to Salesforce’s core artificial-intelligence strategy rather than remaining a separate collaboration product.
Why would Salesforce rely so heavily on Anthropic instead of building every AI layer itself?
The partnership reflects an emerging division of labour within enterprise artificial intelligence.
Anthropic provides frontier model capabilities including reasoning and complex task execution. Salesforce contributes decades of accumulated customer data structures, business rules, permissions, workflow logic and enterprise applications.
Salesforce Chairman and Chief Executive Officer Marc Benioff framed the combination around the idea that probabilistic artificial intelligence alone cannot operate a business reliably, while traditional deterministic software lacks advanced reasoning. Anthropic Chief Executive Officer Dario Amodei similarly argued that Claude becomes more useful when it can work with the commercial context companies already maintain inside Salesforce. Their comments represent the companies’ strategic rationale rather than independently proven outcomes.
The partnership also runs deeper internally than many vendor relationships. Salesforce says Anthropic uses Salesforce as its preferred CRM, while Claude is the default AI model across several Salesforce employee and engineering workflows.
Salesforce has said Slackbot is generating 8.1 million hours of annualised productivity gains internally, more than double the previous quarter. That is a company-calculated productivity measure and should therefore be treated as evidence of Salesforce’s own internal adoption rather than an independently audited economic saving.
What do Salesforce’s latest financial results say about whether Agentforce is becoming material?
The company’s latest results make the artificial-intelligence strategy increasingly measurable.
Agentforce ARR exceeded $1.5 billion in the second quarter of fiscal 2027 and grew more than 240% year over year under Salesforce’s current reporting definition. Combined Agentforce and Data 360 ARR reached nearly $3.9 billion, representing more than 210% year-over-year growth.
Salesforce also reported 7 billion Agentic Work Units delivered across Agentforce and Slack to date, including 3.2 billion during the second quarter, up 97% sequentially. Bookings for premium products combining traditional Sales and Service capabilities with agentic features more than doubled quarter over quarter.
Those figures help explain why the market response was much stronger than the reaction to many earlier Salesforce AI announcements. Investors had spent much of the previous period questioning whether generative AI would weaken established software companies by reducing the importance of traditional applications.
The August results offered the opposite possibility: incumbents owning enterprise data and workflows may be able to monetise AI precisely because intelligent models require secure access to those systems to perform commercially useful tasks.
Salesforce increased its fiscal 2027 revenue guidance to $46.1 billion to $46.4 billion, representing expected growth of approximately 11% to 12%.
Does the 22.6% Salesforce share-price jump mean the AI debate is settled?
No. The August 27 rally was substantial, but it reflected a combination of earnings, guidance, Agentforce growth and renewed confidence in Salesforce’s artificial-intelligence positioning rather than Claudeforce alone.
Salesforce shares closed at $252.05 on August 27 after closing at $205.62 one session earlier, a 22.58% daily gain. They subsequently closed at $256 on August 28, $257.54 on August 31, $258.11 on September 1 and $256.93 on September 2, showing that most of the initial revaluation held through the following sessions.
The market is nevertheless assigning Salesforce a demanding task.
Agentforce must become a meaningful source of incremental revenue rather than merely replacing functionality customers previously purchased through other Salesforce products. Usage growth must convert into durable economics, and Salesforce must prevent powerful external models such as Claude from eventually weakening the value of the underlying application layer.
Claudeforce is therefore both an alliance and a strategic balancing act. Salesforce wants Anthropic’s models to make its platform dramatically more useful without allowing the model layer to become the primary owner of the customer relationship.
If Salesforce succeeds, Slack, Data 360 and Agentforce could become the governed infrastructure underneath AI-driven enterprise work. If the interface shifts too far toward third-party models, Salesforce could find itself supplying the data and workflows while another company controls how employees experience them.
That unresolved tension is what makes Claudeforce more consequential than another enterprise AI product launch.
Discover more from Business-News-Today.com
Subscribe to get the latest posts sent to your email.
