Salesforce employees are expressing concern about Anthropic's expanding footprint inside Slack, the workplace messaging platform that Salesforce acquired in 2021, according to a report from The Information. The worries center on how deeply Claude, Anthropic's AI assistant, can access and interact with internal company conversations, raising questions that many organizations have yet to fully answer as AI tools move into the core of daily work life.
What Is Driving the Unease?
The concerns appear to stem from Anthropic's Claude Tag feature, which allows users to summon Claude directly within Slack channels and threads. Anthropic has been rolling Claude Tag out more broadly, positioning it as a way to bring AI assistance into the flow of team conversations. For some Salesforce employees, however, having an AI model present inside private and semi-private Slack channels feels like a significant shift in how their workplace communications might be monitored or processed.
Key Facts
- The concerns were reported by The Information, citing Salesforce employees familiar with the situation.
- Slack is owned by Salesforce, which acquired it for approximately $27.7 billion in 2021.
- Anthropic's Claude Tag feature lets users invoke Claude AI directly inside Slack messages and channels.
- Anthropic has positioned Claude Tag as a virtual team member, not merely a search or retrieval tool.
- Enterprise AI deployments are increasingly facing employee pushback over data access and privacy boundaries.
The broader question here is one the entire enterprise software industry is grappling with. When an AI assistant is woven into a messaging platform, employees often have limited visibility into what data the model can access, how conversations are stored, and whether their words might be used to improve the underlying AI. Those are not abstract concerns. They are the kinds of questions that HR and legal teams at large companies are now fielding regularly.
Employees want to know exactly what the AI can see, when it's active, and who controls that access. Right now, many companies are moving faster than their internal policies can keep up with.Enterprise technology analyst, paraphrased from industry commentary
Anthropic's Push Into the Enterprise
Anthropic has been making a concerted push into enterprise workflows over the past year. The company framed Claude Tag not simply as a chatbot shortcut but as something closer to a virtual employee embedded in team communication. That framing drew attention when Anthropic first introduced Claude Tag as a virtual Slack employee, and it appears to be generating fresh scrutiny now that the rollout is widening. Anthropic is hardly alone in this strategy. Competitors including Microsoft and Google have moved aggressively to embed AI assistants into productivity suites, but those integrations have also attracted criticism from employees and regulators alike.
The Salesforce situation is a useful case study in how enterprise AI adoption can create internal friction even when the company deploying the tool is also the one that owns the platform. Salesforce is both the host of Slack and, through various partnerships and its own AI initiatives, a company with strong incentives to see AI tools succeed inside its ecosystem. Yet its own employees are pushing back, which suggests the discomfort is genuine and not easily dismissed. Other large deployments, such as Bristol-Myers Squibb rolling Claude out to 30,000 staff, will likely face similar internal questions as AI access expands.
What Comes Next
For Anthropic, navigating employee sentiment at partner companies is part of the challenge of selling into large organizations. A tool that managers champion can still face resistance on the ground if workers feel their communications are less private than before. Clear data governance policies, opt-in controls, and transparent documentation of what the AI can and cannot access tend to reduce friction, but those guardrails take time to build and communicate effectively.
The Salesforce situation is unlikely to derail Anthropic's enterprise ambitions on its own, but it does illustrate that the path from product launch to widespread workplace acceptance is rarely smooth. As AI assistants become standard features of platforms like Slack, companies and their AI vendors will need to invest as much in employee trust as they do in the underlying technology.
