Sovereign AI·Europe

Anthropic Retracts Covert Function Targeting Chinese Users

Global AI Watch · Editorial Team··5 min read
Anthropic Retracts Covert Function Targeting Chinese Users
Editorial Insight

Anthropic's covert monitoring operation marks a significant shift in AI policy scrutiny, affecting cross-border software deployment.

Key Points

  • 1First known instance of covert monitoring by Claude Code.
  • 2Changes user trust dynamics, especially in China.
  • 3Increases scrutiny on Anthropic's compliance with data laws.

What Changed

Anthropic's Claude Code, a programming tool, was found to include a covert monitoring function on July 1, 2026. This hidden feature checked if users were located in China or connected to Chinese domains, without disclosure in its release notes for version 2.1.91. This is the first recorded use of such a monitoring tactic within the Claude ecosystem, prompting significant privacy concerns and discussions around data sovereignty, especially for Chinese users.

Strategic Implications

The revelation shifts the trust dynamics for Anthropic, potentially diminishing brand authority in global markets where data privacy is paramount. While the company claims the function aimed to prevent unauthorized use and distillation by Chinese developers, the breach could result in increased scrutiny and regulatory actions. Additionally, it challenges the balance of power between technology developers and regulators, highlighting the need for transparent software practices.

What Happens Next

Anthropic plans to remove the controversial feature in the next software update, but ongoing regulatory reviews are likely. Chinese authorities may impose stricter compliance checks on foreign AI tools, while Anthropic might bolster its user data policies. By the end of Q3 2026, expect increased transparency in software updates across the industry to restore user confidence.

Second-Order Effects

This incident may increase demand for open-source and transparent software alternatives in regions concerned about data privacy. Additionally, it could lead to a regulatory ripple effect, influencing data protection laws and cross-border AI usage policies, particularly impacting companies reliant on foreign markets like China.

Free Daily Briefing

Top AI intelligence stories delivered each morning.

Subscribe Free →

Explore Trackers