You also seem to have a bug where people get randomly invoiced: https://news.ycombinator.com/item?id=47693679

I got a random invoice for $45.08 back in March, despite not having auto top up enabled. Trying to reach support met with a brick wall. Based on the post I linked to, I'm not the only one facing this problem.

But why did you say that

> I need to let you know that we are unable to issue compensation for degraded service or technical errors that result in incorrect billing routing.

What prevents you from issuing compensations?

Thanks for the follow up here and the transparency.

For those of us not on X, what are the best communication channels for us to follow this sort of communication?

Could really use a post-mortem to set the story straight. The apparently-hallucinated support response copied-pasted by the submitter showing up in the github issue thread is very misleading without scrutiny

"Our support flow wasn't set up"

Would be more accurate. It still isn't setup. Talking to a bot as support who only tells you to talk to the bot for support is not actually support at all. It looks like support, but there's no way to ACTUALLY GET support.

Is it complex? I was somewhat taken aback by how simple it was. Still very confused as to how it could happen.

Have a look at https://github.com/anthropics/claude-code/issues/54497

I can’t use Claude Code online at all

https://xcancel.com/trq212/status/2048495545375990245

Please do explain why someone at Anthropic decided, on purpose, to write code that says something along the lines of: "if ( git_history_str contains "HERMES.md" ... )" then { bill more money }

Somebody (or something) wrote this code. This bug wouldn't be happening for any other reason. It's not a glitch, an oversight, a feature gap, or a temporary outage. It is a piece of written code in your system.

Everyone here is upset about the $200, which is probably much less money than the time that engineer spent ranting about the overcharge on GitHub.

The real problem in my mind is that that bit of code existed in the first place.

Why?

Are you vibe coding your billing!?

Without review!?!?

Or worse, a human being decided to add this to your code base? And nobody noticed or flagged it during code review?

Or much, much worse, Anthropic is purposefully ripping off customers?

This deserves a thorough post-mortem.

I’ve had similar terrible experiences with the Claude support bot when my usage limit was disappearing after a few minutes using Sonnet. I asked for help, asked for escalation, asked for a human, anything. All I got was a non-answers from an AI. I won’t spend real money on Claude now. I’m ok with losing $20 if there’s a rug pull of one way or another, but not $200.

Please, please, please hire more humans with the actual ability to do the right thing for support if your AI agents can’t do the job.

Huh? First off, to have gotten this update when it was posted, I would have needed to:

1) have a Twitter account (which is the virtual equivalent of going to the Nazi bar for a beer, so I don’t)

2) Follow you and be aware that you work for Anthropic

Your support flow is nonexistent, and I hope an acknowledgement/apology/post mortem/etc is forthcoming on your own website, or someplace else that’s, you know, official.

Edit: I’d also like to echo another reply which is flagged for some reason, which points out that

> Our support flow wasn't set up to route a complex bug like this to engineering

Is demonstrably untrue, because an engineer (actually Boris, who is the lead engineer of Claude Code as far as I’m aware) very quickly claims to have fixed the bug four days ago and then ignored all of the follow up comments regarding the refund. From the outside, it seems like maybe the inverse of what you said is more accurate: your engineers aren’t able or willing to route issues like this to support/billing to be able to issue refunds.

[dead]