Salesforce Acquires Fin for $3.6B to Fuel AI Customer Service

On June 15th, Salesforce announced that it was acquiring Fin, an AI-powered customer service tool, for $3.6 billion. Fin is the company formerly known as Intercom, which has been operating for 15 years but recently changed its name to reflect its focus on an AI-powered agent. This acquisition is slated to be finalized in the first quarter of 2027, during Salesforce’s fourth quarter of its fiscal year 2027.

Fin offers a simple value proposition: provide an AI agent that can answer customer questions over a variety of different communication channels (such as live chat, WhatsApp, SMS, phone calls, and Slack) without having to escalate issues to a human agent. They recently released their own Apex model and have developed an internal agent called Operator. According to Salesforce, they hope to leverage Fin’s technology and talent to improve their existing product, Agentforce, which allows companies to build custom AI agents for automating various tasks. After the acquisition, Fin co-founder and CEO Eoghan McCabe will remain at the helm.

When discussing the acquisition, Salesforce CEO Marc Benioff emphasized the importance of capabilities, stating that “with Fin, we gain proven agent technology, a deep commitment to customer success, and an amazing AI team.” In other words, Salesforce believes that it is easier to acquire an existing customer service agent with users rather than developing one themselves.

1Why a support deal matters to you

One of the first areas where AI is being applied is in customer service, since it is a high-cost, repetitive, and quantifiable task. By paying $3.6 billion for Fin, Salesforce is sending a clear message that they believe that most first-level customer service interactions will be handled by AI agents rather than humans in the coming years. When you call your bank, telecom provider, or online retailer, there’s a good chance that you’ll be speaking to a model instead of a human, and these models can now handle complex tasks like processing returns, invoicing, and modifying user accounts without any need for human intervention.

This has both positive and negative implications. On the one hand, AI-powered customer service can be quick and responsive, even at odd hours. On the other hand, poorly designed AI agents can create frustrating barriers between users and actual humans, refusing to transfer you to a human representative if they cannot resolve your issue. These decisions are being made today, and they are driven by financial considerations rather than user experience.

In my opinion, the technology is legitimate and can frequently outperform human-powered call centers. However, the danger lies not in the intelligence of the agent, but in its potential use as a barrier to prevent users from speaking to a human representative when they receive a “no” response.