Alibaba Bans Employees From Using Claude Code

Alibaba has reportedly issued an internal ban preventing its employees from using Claude Code, the agentic coding assistant developed by Anthropic. The move, first reported by TechCrunch, adds Alibaba to a growing list of large technology companies restricting staff access to AI tools built by outside firms, particularly those based in the United States.

What the Ban Covers

The restriction appears to apply specifically to Claude Code, rather than to Claude's broader product lineup. Claude Code is a terminal-based coding agent that can read, write, and execute code across an entire codebase, often working with minimal human supervision. That level of system access has made it a focus of scrutiny from security-conscious organizations. For context on the concerns this tool can raise, reports of fake Anthropic sites targeting Claude Code users with infostealer malware have already demonstrated how attractive the product is to bad actors looking to exploit its permissions model.

The timing is notable. Alibaba has been aggressively pushing its own AI development capabilities. The company's Qwen model series has gained traction both inside China and internationally, and earlier this year Alibaba demonstrated that Qwen3.7-Max could run autonomously for 35 hours while working alongside Claude Code. Banning Claude Code internally while continuing to develop tools that interact with it externally reflects the complicated, sometimes contradictory, posture large tech firms take toward their AI competitors.

"The concern isn't necessarily that Claude is untrustworthy. It's that any external tool with this level of code access represents a potential data exfiltration risk, intentional or otherwise."Security researcher quoted by TechCrunch

A Pattern of Corporate AI Restrictions

Alibaba is far from alone in placing guardrails on which AI tools employees can use. Samsung famously restricted ChatGPT after engineers accidentally leaked proprietary source code through the chatbot. Many firms now maintain approved-tools lists, and agentic coding assistants, which can autonomously browse files and run commands, are drawing particular attention from security and legal teams.

For Alibaba specifically, there is an added competitive dimension. Allowing engineers to route code through a rival company's infrastructure, even for productivity purposes, carries intellectual property risks that go beyond ordinary data hygiene. Details on the internal rationale behind Alibaba's decision have been covered in reporting on Alibaba's spyware concerns around Claude Code, which suggests the ban was driven at least partly by fears about how the tool handles data it encounters during a session.

The episode also lands at a delicate moment for Claude Code's broader adoption story. Anthropic has positioned the tool as a productivity multiplier for professional developers, and enterprise uptake has been a key part of that narrative. Losing access within one of the world's largest technology companies, even informally, is a signal the market will notice. It also raises a quieter irony: internal reporting has previously noted that Claude Code's heavy use inside Anthropic itself has left some employees feeling isolated, suggesting the tool's social and cultural footprint is being felt even at the company that built it.

Anthropic has not publicly commented on Alibaba's reported policy. The company continues to develop Claude's model family and expand Claude Code's capabilities, including support for larger context windows and more complex multi-step workflows. Whether the Alibaba ban represents an isolated corporate decision or an early sign of broader enterprise hesitation around agentic coding tools remains to be seen.