The model for Sift (formerly Sift Science) was to essentially pull all the data everyone provided in together so anyone using their system benefitted from every

The model for Sift (formerly Sift Science) was to essentially pull all the data everyone provided in together so anyone using their system benefitted from everyone else's data, but without exposing the actual PII of the fraudsters (or people the fraudsters are pretending to be). There are public suspicious IP address lists/lookups but nowadays fraudsters with any level of sophistication tend to get a fresh IP each time. It's a hard problem. You have to decide who you can trust to submit "fraudsters" data and to what extent you trust what they provide. If you start looking at PII (emails, phone numbers, faces, etc. as you say) then even good actors may inadvertently add data of innocent people that fraudsters are fronting as.