Pixelfed (and Mastodon) are increasingly targeted by spammers.
Our current AutoSpam detection using Naive Bayes isn't enough*, so we are building a better detection and mitigation system.
Not only will it leverage (optional) established 3rd party services like SpamHaus, but also threat intel from other trusted "reporter" instances like Pixelfed.social.
It's called https://FediThreat.net, and will be available soon!