Very lengthy process; Initial recruiter call, then manager round, then take home assignment, then 4 subsequent rounds after that.
Its a very very impersonal process, they're so obsessed about getting the safety component of AI right that they very visibly seem to have completely forgotten about the "human" side of AI. They run the recruitment process like AI would if it was given autonomy over the entire recruitment process.
They demand an extraordinary amount of effort (specifically the take home assignment), but explicitly tell you no feedback will be given. Even after such a gruelling, 8-9 round total process, they don't give you any feedback or suggest areas where you can improve.
It's a very inhumane and unempathetic company. They are no doubt the leading AI Lab at the moment, and they are also the safest, but their obsession over safety and misuse of AI has made them deal with the market (labour and otherwise) as a safety obsessed AI agent would.
I wouldn't bother applying, OpenAI is a much more humane process and they pay 60-70% better than Anthropic does as well.