Story perspectives

Alibaba Bans Anthropic Claude, Cites Fake Account Abuse

7/4/2026

Summary

Alibaba Bans Claude

Alibaba will ban employees from using Anthropic’s Claude Code starting July 10.
The company orders staff to delete the Claude series (Sonnet, Opus, Fable) and switch to Qoder.
Alibaba deems Claude Code high-risk because it can read timezone and proxy data.
Anthropic alleges Alibaba used 25,000 fake accounts from April 22 to June 5 to generate 28 million Claude conversations and sent a July 10 letter to the U.S. Senate Banking Committee that prompted the ban.

Summary

Banks Restrict Claude

JPMorgan and Goldman Sachs restricted employee use of Anthropic’s Claude models in Hong Kong in June, citing Greater China licensing limits.
Alibaba’s Hong Kong shares slipped 0.7% after the ban, underperforming the Hang Seng Index, which rose 1.3% that session.
A U.S. Senate Banking Committee hearing on AI is set after the June 10 letter, echoing the White House’s April claim that China stole U.S. AI laboratory IP on a scale.

History

Historical Context

In October 2019 the U.S. Commerce Department added SenseTime to the Entity List over alleged Xinjiang surveillance, halting its U.S. technology access and prompting Alibaba to reassess AI contracts. The action echoed China’s May 2019 “unreliable entities” response and followed SenseTime’s 2018 designation as a state-backed AI champion, a status reflected by its 54 million GPU cores.

1 of 3

Alibaba Bans Claude

Alibaba will ban employees from using Anthropic’s Claude Code starting July 10.
The company orders staff to delete the Claude series (Sonnet, Opus, Fable) and switch to Qoder.
Alibaba deems Claude Code high-risk because it can read timezone and proxy data.
Anthropic alleges Alibaba used 25,000 fake accounts from April 22 to June 5 to generate 28 million Claude conversations and sent a July 10 letter to the U.S. Senate Banking Committee that prompted the ban.

1 / 3