130
Claude Code sends your email address to websites
(github.com)
Yet another reason to simply not use "AI" "tools".
I've had a ton of success talking people out of considering AI as an option by calling it a "probability based automation". I forget where I heard the term originally or I'd link the article.
That's a good one! I'm gonna use it :3
IDK man. You clearly didn’t read what actually happened so you are clearly okay with off loading some of that internal compute.
I actually read the entire thing, and stand by what I said. Would you like to attempt a proper argument, by proving what I said to be baseless?
I hate AI more than anyone, but it's vibe coders who are being stupid here: https://news.ycombinator.com/item?id=50387577
The SEC API documentation requires your email in the User-Agent
It reminds me of the idiots who had their OS trashed because the AI thing was not sandboxed and they had no backup. Well, duh, we told you a billion times already.
Thanks for the link. This one would be funny if someone did when AI agent accessing a web server is detected lol: https://news.ycombinator.com/item?id=50388101
Someone might call this Karma…
I receive dozens of scam emails per day because claude code[...]
Oh no! Anyway...
TL;DR from what I can tell:
They loaded the user email address into the model context in Claude Code, because users wanted it to correctly generate that in certain places, like when documenting that it was the user who made a certain change.
But webpages can ask for the e-mail address to be sent along with the user agent and so the model just, well, did that.
Man the comments are a fucking shit show lol
Glad i dont use Claude.
this post was submitted on 06 Nov 2026
130 points (99.2% liked)
Privacy
11177 readers
189 users here now
A community for Lemmy users interested in privacy
Rules:
- Be civil
- No spam posting
- Keep posts on-topic
- No trolling
founded 3 years ago
MODERATORS