THE APEX TIMES
Salesforce and Anthropic bring Claude into the CRM workflow with a new “Salesforce in Claude” pilot
The Claudeforce partnership launches a Salesforce plugin for Claude with 37 prebuilt sales skills, designed to reason over live revenue data and execute actions inside existing Salesforce permissions and rules. A broader rollout is slated for September 2026.
Salesforce is rolling out a new way for sales teams to use AI, positioning Claude from Anthropic as an operating layer over CRM data and workflows rather than another standalone chatbot. In a company update published Monday, Salesforce said its “Claudeforce” partnership is delivering “Salesforce in Claude,” a plugin that brings Salesforce’s sales and revenue-cycle capabilities directly into Claude for guided reasoning and governed actions.
Salesforce says the product is designed to work with the permissions and business rules a company already runs on, so Claude can only see and act on information a given user is authorized to access. The company frames the shift as moving beyond software “screens” and fixed interfaces toward AI that can reason across systems and then take the next step. Salesforce also claims the approach reduces the time sales reps spend stitching together account history, pipeline context, and prior communications into a workable plan.
The launch begins with a sales-focused plugin that ships with 37 prebuilt “skills” built for seller workflows. Salesforce describes the skills as engineered to draw on Claude’s reasoning and agentic tool use, and to avoid what it calls “generic CRM prompts wrapped around an API.” The initial set is aimed at tasks across the sales motion, including onboarding that reads a seller’s enterprise context and generates a tailored interactive dashboard, plus pipeline updates and deal preparation use cases.
Salesforce says the product can surface what needs attention “today” by analyzing live pipeline changes and producing an action plan, including what moved overnight and what is at risk. It also describes a deal Q&A experience that goes beyond a health score, generating objections likely to arise and a plan to address them based on historical record changes. For account growth, Salesforce says Claude can identify expansions or renewals embedded in accounts and turn top prospects into outreach-ready material.
On action-taking, Salesforce says it routes every step through Salesforce, so company-specific rules are enforced when AI updates records or runs multi-step workflows. The company adds a control layer for outbound communications, describing an option for Claude to check with a user before sending messages outside the company. It also says record edits are limited to the specific fields the AI intends to update, with humans directing and agents executing under Salesforce governance.
Underpinning the plug-in is Salesforce’s “Headless 360” concept, which the company describes as exposing data, apps, workflows, agents, and governance as capabilities that AI can call over a secure Model Context Protocol (MCP) connection. Salesforce says this design is meant to avoid UI-bound access, enabling an AI agent to use the same underlying capabilities wherever it is working, including inside Claude and also in Slack, using a consistent set of data and answers.
Salesforce also ties the new offering to prior work with Anthropic, including “Claude Tag” in Slack and the rollout of Claude-related capabilities in Agentforce, as well as Anthropic being a launch partner for “Slack Code,” an environment where engineering teams collaborate on code in the open. The company describes Claudeforce as an effort to put AI into tools people already use across roles, from engineers to knowledge workers, while treating Salesforce as the “trusted” system of record for business context.
For rollout, Salesforce said the product is available today to select pilot customers and is planned for an open beta starting in September 2026. The initial release is described as delivering prebuilt skills for Sales first, with additional skills for other teams expected in the months after launch. Salesforce also said it is using the offering internally, describing itself as a “Customer Zero,” with pilots run on real pipeline, accounts, and deals.
Why It Matters
- The announcement indicates a competitive push to move AI from assistant-style Q&A toward workflow execution in enterprise systems, using governance and permissions as guardrails.
- If widely adopted, a CRM-in-AI approach could change how sellers prepare for calls and manage pipeline, with less manual navigation across accounts and history.
- The emphasis on routing actions through existing CRM rules suggests a path to reduce compliance and data-access risk that often comes with AI tools.
- A September open beta may provide early evidence of demand from revenue teams for AI experiences embedded in the tools they already use, rather than separate applications.
Sources
Key Facts
- Salesforce and Anthropic are partnering under “Claudeforce” to connect Claude with Salesforce CRM data and workflows through “Salesforce in Claude.”
- Salesforce in Claude is described as a plugin for Claude that launches with 37 prebuilt sales-focused skills for tasks such as pipeline review, deal preparation, and outreach.
- Salesforce says Claude’s answers and actions are governed by existing Salesforce permissions and business rules, with controls over whether and when messages are sent externally.
- The company says the product is available to select pilot customers now and planned for an open beta starting in September 2026.
- Salesforce says the architecture is built on “Headless 360,” exposing Salesforce capabilities to AI through MCP with no custom integration work described for the end user.
Technology Related
NVIDIA heads into Q2 2027 earnings watch as markets price in “whisper” expectations
Ahead of the company’s next quarterly results, investors and analysts are closely tracking a chorus of pre-announcement estimates and question marks around how the latest demand and product momentum will show up in the numbers.
Nvidia reports revenue that nearly reaches $100 billion for the quarter, while investors weigh the next proof point
The AI chip leader says its latest quarter delivered a major sales surge and outperformed Wall Street expectations, but the stock’s muted reaction outlines that the bar for durability remains high.
Nvidia pushes back on claims its AI investments are “circular financing,” as scrutiny grows
In remarks highlighted by Yahoo Finance, Nvidia disputed characterizations of its AI-related partnerships and investments as a closed-loop funding strategy, even as the chipmaker benefits from surging demand for data-center infrastructure.
Stocks Edge Lower as Inflation Fears Rise and Investors Await Nvidia Earnings
A hotter-than-expected July inflation read and higher bond yields shifted expectations for rate cuts, nudging equities down and pressuring chip-related names ahead of Nvidia’s earnings. The market also digested a new ruling involving Meta.
Nvidia reports broad Q2 strength, lifting FY28 revenue outlook and pushing shares higher after hours
Jensen Huang tied the results to continued momentum in AI infrastructure, including production of the Vera Rubin system.
Salesforce shares jump after hours as earnings top expectations and company lifts outlook
Salesforce reported adjusted Q2 results that beat Wall Street forecasts, then raised guidance and expanded a partnership tied to Anthropic’s Claude. The stock rose sharply in after-hours trading.
Meta settlement with U.S. states raises new pressure across social media, industry observers say
A widely reported settlement between Meta and dozens of U.S. states is being framed as a potential turning point, with analysts pointing to the likelihood of further regulatory and legal scrutiny aimed at social platforms.
Nvidia reports $96.2 billion quarterly revenue as AI demand pushes forecasts higher
The graphics and AI-chip leader said quarterly revenue more than doubled from a year earlier and raised its outlook, reinforcing that customers are spending more aggressively on artificial intelligence infrastructure.
Meta shares rise after reported $18 billion child-safety settlement with U.S. states
A reported settlement covering child-safety issues with U.S. states appeared to reduce a major legal overhang for Meta, helping lift the company’s stock on Aug. 26.
Amazon Web Services and Nvidia expand AI infrastructure deal, adding 2 million GPUs
AWS and Nvidia said they are increasing capacity for AI workloads by planning the deployment of 2 million additional Nvidia GPUs, indicating continued demand for accelerated compute as companies build and run large-scale machine learning systems.