Salesforce agreed on Monday to acquire Fin, the AI customer-agent company formerly known as Intercom, for about $3.6 billion, a deal that underscores how aggressively enterprise software giants are buying their way to the front of the agentic AI race. As reported by CNBC, Fin’s technology will complement Salesforce’s own AI agent product, Agentforce, deepening the company’s push to sell autonomous digital workers to large enterprises. The transaction is expected to close in Salesforce’s fiscal 2027 fourth quarter.

The acquisition is a clear signal about where the software industry believes the next wave of value will be created. For the better part of two decades, business software was about systems of record: databases that stored customer information, sales pipelines, and support tickets. The emerging battle is over systems of action, software that does not merely store information but autonomously resolves the work, answering the customer, closing the ticket, completing the task. Fin is one of the more credible independent players in that shift, and Salesforce paid a premium to bring it in-house rather than let a rival do so.

What Salesforce Is Buying

Fin’s flagship product, called AI Agent, handles customer queries across a wide range of channels, including chat, email, WhatsApp, text message, phone, and Slack. What distinguishes it from a generic chatbot is the engine underneath. The AI Agent is powered by Fin’s proprietary model, Apex, which the company built specifically for customer support and which it says delivers industry-leading resolution rates that outperform top commercially available frontier models on support tasks.

That last claim is the crux of the deal’s logic. General-purpose large language models are extraordinary generalists, but enterprise customer support is a narrow, high-stakes domain where accuracy, tone, and the ability to actually resolve an issue matter more than broad knowledge. A model purpose-built and tuned for support, with access to a company’s own knowledge base and the ability to take actions on the customer’s behalf, can outperform a bigger but less specialized model on the metric enterprises care about most: how many customer problems get solved without a human stepping in. By acquiring Apex along with Fin’s product and team, Salesforce gets a domain-tuned model rather than having to build one from scratch.

Fin’s lineage adds to the story. The company was previously known as Intercom, a well-established name in business messaging and customer communication. That history means Fin arrives with real enterprise relationships, a mature product, and engineering depth, not just a promising demo. For Salesforce, which serves a vast base of large customers running support operations, the cultural and product fit is direct.

Strengthening Agentforce

The strategic centerpiece is Agentforce, Salesforce’s platform for building and deploying AI agents inside the enterprise. Salesforce has positioned Agentforce as the way its customers move from AI that assists humans to AI that completes work on its own, and the company has tied a meaningful part of its growth narrative to the platform’s adoption. Folding Fin’s purpose-built support agent and the Apex model into Agentforce gives Salesforce a stronger, more proven offering precisely in customer service, one of the highest-volume and most measurable use cases for agentic AI.

The competitive context explains the urgency. Technology firms are racing to roll out usage-based autonomous digital workers across enterprises, and the pricing model itself is shifting. Where traditional software charged per seat, agentic AI is increasingly sold per outcome or per action, which aligns vendor revenue with the work the software actually performs. That model rewards whoever can demonstrate the highest resolution and reliability, because customers pay for results. Owning a model with leading support-resolution rates is therefore not a vanity acquisition; it is a direct input into the metric that determines revenue.

Customer service is an especially attractive beachhead for agentic AI because the economics are so legible. Support organizations are large, costly, and metric-driven, with established measures like first-contact resolution, average handle time, and customer satisfaction that translate cleanly into dollars. Every query an AI agent resolves correctly without a human is a quantifiable saving, and every escalation it prevents protects a customer relationship. That makes support the rare enterprise function where the return on autonomous AI can be demonstrated quickly and defended in a budget meeting, which is exactly why Salesforce, Microsoft, ServiceNow, and a wave of startups are all converging on it. Owning a model that wins on resolution rate, as Fin claims Apex does, is a direct claim on that budget.

This is the same agentic wave reshaping other corners of business technology. We covered an early consumer-facing example in our reporting on Robinhood’s AI agent for agentic trading and credit, where autonomous software takes actions on a user’s behalf rather than merely surfacing information. Salesforce’s deal is the enterprise-software version of the same idea, applied to the back office and the contact center.

The Bigger Picture on Enterprise AI Spending

The Fin acquisition fits a broader pattern of enormous capital flowing toward AI capability. Across the industry, the largest technology companies are spending at a scale that would have looked reckless a few years ago, convinced that the firms which establish leadership in AI now will compound that advantage for years. We detailed the infrastructure side of that build-out in our coverage of Nvidia’s $40 billion in AI investments across its supply chain, and the capital-allocation debate it raises in our analysis of big tech AI spending versus investor payouts.

Salesforce’s $3.6 billion check sits at the application layer of that stack rather than the silicon layer, but it reflects the same conviction: that demand for AI that can autonomously do enterprise work is real and durable, and that the window to assemble a leading position is now. Buying Fin lets Salesforce skip the long, uncertain path of building a best-in-class support model internally and instead acquire one with a track record, a customer base, and a team that has already shipped.

For investors, the deal is best read as both offensive and defensive. Offensively, it strengthens the most strategically important product in Salesforce’s portfolio at a moment when agentic AI is the company’s clearest growth story. Defensively, it denies a proven asset to competitors who are pursuing the same enterprise customers with their own agent platforms. In a market where the difference between leading and trailing in AI can be measured in resolution rates and customer trust, paying up for a domain-tuned model and an experienced team is a calculated move rather than an extravagant one.

What to Watch

The questions now turn to execution. Salesforce must integrate Fin’s Apex model and product into Agentforce without disrupting either Fin’s existing customers or its own, and it must convince enterprises that an acquired agent and a homegrown platform will work as one coherent system. The deal’s close in Salesforce’s fiscal 2027 fourth quarter gives a timeline, but the real test is adoption: whether large customers route more of their support volume through Agentforce-plus-Fin and whether the combined product sustains the resolution-rate edge that justified the price.

If it works, Salesforce will have meaningfully extended its lead in one of the most valuable AI use cases in the enterprise, customer service at scale. If integration falters or rivals close the resolution-quality gap, the acquisition will still have bought Salesforce time and talent in a race where both are scarce. Either way, the $3.6 billion deal is another marker of how completely agentic AI has moved from pitch decks to balance sheets in 2026.

How much did Salesforce pay for Fin?

Salesforce agreed to acquire Fin, the AI customer-agent company formerly known as Intercom, for about $3.6 billion. The deal is expected to close in Salesforce’s fiscal 2027 fourth quarter.

What does Fin's AI Agent do?

Fin’s flagship product, AI Agent, handles customer queries across chat, email, WhatsApp, text message, phone, and Slack. It is powered by Fin’s proprietary model, Apex, which is purpose-built for customer support and claims industry-leading resolution rates that outperform top commercially available frontier models on support tasks.

How does the deal affect Salesforce's Agentforce platform?

Fin’s purpose-built support agent and Apex model will complement Agentforce, Salesforce’s platform for building and deploying AI agents. The acquisition strengthens Agentforce in customer service, one of the highest-volume and most measurable use cases for autonomous AI.

What was Fin previously known as?

Fin was previously known as Intercom, an established name in business messaging and customer communication. That history means it brings mature enterprise relationships, a proven product, and engineering depth to Salesforce.

Why are software companies racing to acquire AI agent makers?

Enterprise software is shifting from systems of record that store data to systems of action that autonomously complete work. Agentic AI is increasingly sold per outcome or action rather than per seat, which rewards whoever can demonstrate the highest resolution and reliability, making proven AI agents strategically valuable.

When will the Salesforce-Fin acquisition close?

The transaction is expected to close in Salesforce’s fiscal 2027 fourth quarter, subject to customary closing conditions.