I'm sick of @AnthropicAI's shenanigans at this point.

I had reservations about using it in prod since autumn, I don't remember why exactly but it was about something they have done back then.

Now I'm moving from "it's bad" to "it's an actual risk using their models".