China Flags Anthropic’s Claude Code as Backdoor Risk Over Data Transmission Claims
China’s cybersecurity platform flagged a "backdoor" security risk in Anthropic’s Claude Code, alleging the coding assistant can transmit user location and identity data without consent.
Anthropic dismissed the warning, characterizing the functionality as an experimental anti-abuse mechanism and noting the service is not officially permitted in China.
From the sources (8 posts)
@wesrothAlibaba has reportedly banned employees from using Claude Code at work after Anthropic’s coding tool came under scrutiny for features that can identify China-linked users. Employees were told to use Alibaba’s own coding platform, Qoder, in
@financialjuiceAnthropic sent a letter to lawmakers on Alibaba copying AI - NYT.
@firstsquawkANTHROPIC SENT LETTER TO LAWMAKERS ON ALIBABA COPYING AI: NYT
@deitaoneANTHROPIC TOLD LAWMAKERS ALIBABA COPIED AI: NYT
@zerohedge*ANTHROPIC SENT LETTER TO LAWMAKERS ON ALIBABA COPYING AI: NYT
@exec_sumBREAKING: Alibaba banned employees from using Anthropic's coding tool to prevent Chinese identities being exposed This comes days after Anthropic sent a letter to Congress members warning that Alibaba was secretly copying its AI technolog
@reutersChina issues 'backdoor' security alert over Anthropic's Claude Code
@cointelegraph🇨🇳 NEW: China's cybersecurity platform flags a "backdoor" security risk in Anthropic's Claude Code, claiming it can transmit user location and identity data without consent. Anthropic says the flagged feature was an experimental anti-abuse