Possible solutions: - mCaptcha https://github.com/mCaptcha/mCaptcha - Google's invisible recaptcha this would get triggered randomly on users based on a variable (maybe specific to a user?) google's invisible recaptcha would be the most non-UX-invasive but obviously privacy invasive mcatpcha would be slightly UX invasive as it prompts the user because of the proof-of-work it uses, which going this route would be privacy respecting mcaptcha also might not stop any botters due to it just using more energy/processing power to run those checks could have a way to switch between them on the fly? or maybe per-user moderation actions to switch between the two