> So, just curious, how does filtering out spam work, if the email provider does not have access to the contents? Much (most?) of spam filtering is done using

> So, just curious, how does filtering out spam work, if the email provider does not have access to the contents? Much (most?) of spam filtering is done using DKIM/SPF/DMARC/ARC, domain reputation, and IP reputation, none of which require access to the contents. I'm not saying that you should go and make all your email subjects "Free herbal viagra" or whatever, but modern spam filtering algorithms revolve somewhat less around those kinds of trigger phrases than they used to. Also, spam is by definition sent in bulk, so as long as the emails are hitting either some accounts on other email providers or else are hitting some honeypots on ProtonMail then you still get the benefit of being able to have the sender blacklisted based on the contents.