Alibaba bans Claude Code as a 'backdoor.' An ex-Anthropic engineer says that's not what a backdoor is
Reuters, one anonymous source, one loaded word: Alibaba is pulling Claude Code off employee machines over 'backdoor risks' and steering them to its own Qoder. In the thread, someone who worked inside Anthropic's orbit reaches for the definition first.
The alleged 'backdoor' is a recently shipped Claude Code change that reads the machine's timezone and locale and phones a bit of location data home. Alibaba's reported response: ban the tool company-wide and redirect developers to its in-house coding agent, Qoder.
A contractor who was the first engineer at Stainless, later bought by Anthropic, cut in with the pedantic-but-fatal objection: telemetry that reports your timezone is not a backdoor. And the same firms flagging the risk already host their entire private codebase on GitHub.
What Reuters actually reported — and how thin the sourcing is
The story is a single-source Reuters brief (mirror on archive.is, since Reuters blocks direct reads): Alibaba is banning Claude Code in the workplace over alleged backdoor risks, "a source says," and is pointing staff at its own coding platform, Qoder. No named official, no forensic write-up, one word doing all the work: backdoor.
The underlying behavior is real and much smaller than the headline. A few days earlier, a separate HN thread dissected a Claude Code build that detects a Chinese timezone/locale and transmits some location data. That is the "backdoor."
"That's not what a backdoor is"
The sharpest correction in the thread came from someone with a foot inside the vendor's world.
"That's not what a backdoor is…" — a Hamburg contractor who was the first engineer at Stainless, acquired by Anthropic
A backdoor is covert remote access. Locale telemetry that leaks a timezone marker is a privacy smell, not a way in. One commenter deflated the stakes further:
"If you think POTUS would lose sleep about a date format timezone marker, I don't know what to tell you." — a commenter sizing the alleged severity against the rhetoric
The trust math the ban ignores
The cleaner puncture is about consistency. These companies already ship their crown jewels to American infrastructure every day.
"They trust GitHub not to read their code but they don't trust an AI from Microsoft not to read it? It would be schizophrenia" — an engineer pointing at where the code already lives
A sub-thread tried to spin the ban as evidence Alibaba had finished distilling Anthropic's model and no longer needed it — and got debunked on the timeline, then corrected again for confusing Alibaba with the separate lab that ships GLM. What survives all of it is the read that reframes the whole story: this looks less like a security discovery and more like a PR battle between two firms, fought with a scary noun.
Get breakdowns like this every day
The top Hacker News stories, unpacked — by email, Telegram, or RSS.
He built a product on 'transcripts are the new oil.' Now he reports zero benefit and recants
Amol Kapoor once believed an agent's saved session logs were more valuable than the code they produced. His team measured it across hundreds of sessions and found nothing — so he wrote the post killing his own idea, then had to defend it live against a practitioner who says the opposite works for him.