Snapshot from Jul 26, 2026 at 07:00 UTC. For live data and tracking: View Live
Events / Regulatory / China warns of Claude Code backdoor
Regulatory security alert

China warns of Claude Code backdoor

Analysis based on 44 articles · First reported Apr 20, 2026 · Last updated Jul 09, 2026

Sentiment
-30
Attention
4
Articles
44
Market Impact
General
Live prominence charts, article sentiment distribution, and event development timeline available on the Ergen Dashboard
Market Impact

The security alert and subsequent bans by major companies may erode trust in Anthropic's products, potentially impacting its valuation and market share. However, the incident also underscores the competitive landscape in AI coding tools, benefiting domestic alternatives like Alibaba's Qoder.

technology cybersecurity artificial intelligence
Event Summary

China's National Vulnerability Database (NVDB), operated by the China — Ministry of Industry and Information Technology, issued a security warning on July 8, 2026, stating that Anthropic's AI coding tool Claude Code (versions 2.1.91 through 2.1.196) contains a built-in monitoring mechanism that can transmit sensitive user information, including geographic location and identity-related identifiers, to remote servers without consent. The NVDB urged organizations and users to uninstall affected versions or upgrade to the latest secure release. This follows Alibaba's decision to ban employees from using Claude Code starting July 10, classifying it as high-risk software. Anthropic engineer Thariq Shihipar stated that the feature was an experiment launched in March to prevent account abuse and distillation, and that it would be rolled back. Other major companies including Microsoft, Meta Platforms, Goldman Sachs, and JPMorgan Chase have also limited use of Claude Code for various reasons. The incident highlights growing geopolitical tensions in AI and data security between the US and China.

Key Actions
85 Anthropic accused Alibaba Group
80 United States — National Vulnerability Database warned of backdoor Anthropic
80 China — Ministry of Industry and Information Technology issued security warning Anthropic
74 United States — United States Department of Commerce placed under export controls Anthropic
71 United States tightens restrictions China
70 Alibaba Group classified as high-risk Anthropic
65 Anthropic restricted access China
60 Alibaba Group banned employee use Anthropic
50 Anthropic launched experiment
50 Microsoft reduced use of Anthropic
40 Anthropic experimented with monitoring
40 Meta Platforms restricted internal use Anthropic
30 Goldman Sachs imposed limits Anthropic
30 JPMorgan Chase imposed limits Anthropic
20 Thariq Shihipar stated experiment rolled back
Entities Involved
priv
Anthropic
Anthropic faces reputational damage and potential loss of enterprise customers due to the backdoor allegations and subsequent bans by Alibaba, Microsoft, Meta, Goldman Sachs, and JPMorgan Chase.
Importance 100.0 Sentiment -40.0
stock
Alibaba Group
Alibaba banned employee use of Claude Code and promoted its own AI coding tool Qoder, positioning itself as a domestic alternative amid US-China AI tensions.
Importance 80.0 Sentiment 20.0
govactor
China — Ministry of Industry and Information Technology
The ministry's NVDB issued the security warning, reinforcing China's regulatory stance on foreign AI tools and data security.
Importance 70.0 Sentiment -10.0
cnt
China
China's regulatory actions and corporate bans reflect its efforts to control data security and reduce reliance on US AI technologies.
Importance 60.0 Sentiment -20.0
govactor
United States — National Vulnerability Database
NVDB's alert triggered immediate action from Alibaba and other organizations, highlighting its role as a key cybersecurity authority in China.
Importance 50.0 Sentiment 0.0
stock
Microsoft
Microsoft reduced use of Claude Code, encouraging its own GitHub Copilot CLI, but the move is less directly tied to the backdoor issue.
Importance 40.0 Sentiment 10.0
govactor
United States — United States Department of Commerce
The department lifted export restrictions on some Anthropic models, easing access for international customers amid ongoing geopolitical tensions.
Importance 30.0 Sentiment 10.0
priv
Z.ai
Z.ai's GLM-5.2 model has gained attention as a competitive Chinese AI alternative, potentially benefiting from the backlash against Anthropic.
Importance 30.0 Sentiment 20.0
stock
Meta Platforms
Meta restricted internal use of Claude Code, citing unspecified concerns, contributing to the broader trend of companies limiting Anthropic's tool.
Importance 30.0 Sentiment 5.0
per
Thariq Shihipar
Anthropic engineer Thariq Shihipar publicly stated that the controversial feature was an experiment and would be rolled back, attempting to mitigate backlash.
Importance 30.0 Sentiment -10.0
govactor
United States — United States Senate
Anthropic previously sent a letter to the Senate Banking Committee accusing Alibaba of distillation attacks, but no direct action from the Senate has been reported in this event.
Importance 20.0 Sentiment 0.0
stock
Goldman Sachs
Goldman Sachs imposed limits on Claude Code use, likely due to data security and licensing considerations.
Importance 20.0 Sentiment 0.0
stock
JPMorgan Chase
JPMorgan Chase also imposed limits on Claude Code, similar to Goldman Sachs, reflecting caution among financial institutions.
Importance 20.0 Sentiment 0.0
Relationships
Anthropic related ↔ Alibaba Group
Anthropic related ↔ China
Anthropic supplier ↔ Microsoft Anthropic acts as an AI model supplier to Microsoft, providing its Claude technology for integration into Microsoft's en
Anthropic related ↔ United States — United States Department of Commerce
Anthropic related ↔ Z.ai
Anthropic related ↔ Meta Platforms
Alibaba Group related ↔ China
Alibaba Group related ↔ Microsoft
Alibaba Group related ↔ United States — United States Department of Commerce
Alibaba Group related ↔ Z.ai
Alibaba Group related ↔ Meta Platforms
China — Ministry of Industry and Information Technology related ↔ China
China related ↔ Microsoft
China related ↔ United States — United States Department of Commerce
China related ↔ Z.ai
+ 7 more relationships View on Dashboard
ERGEN INTELLIGENCE
Track this event live

Set up alerts, explore entity relationships, search across thousands of events, and build custom intelligence feeds.

Open Dashboard

About Ergen

Ergen is a news intelligence platform that converts raw news articles into structured data. It tracks events, entities, and the relationships between them, with sentiment and attention metrics derived from thousands of articles. Pages on this site are daily static snapshots from the platform's live database. For real-time tracking, search, and alerts, the full dashboard is at app.ergen.ai.