[dns_snek]

> Multi-stage moderation - Plausibility signals (AI filtering, account activity) - Ability to "reply" or mark as "challenged"

Will you require hard evidence as a matter of policy before publishing any allegations? That's the only thing that would set you apart from many previous implementations of this idea which always seem to turn into hateful and defamatory platforms.

"Plausibility signals" like AI filtering are only good for filtering outright spam and bot activity, they're useless for determining truthfulness of claims being made. If your goal is to publish truthful information then please take some lessons from the legal system because truth can only be ascertained by analyzing evidence, not by evaluating the accuser's social standing ("account activity") and asking a random village idiot for their opinion on one side of the story ("AI filtering").

And the whole process of "people can respond and ask for it to be taken down" is not good enough because by that time the damage could already be done. People who have done nothing wrong have no reason to proactively monitor and "curate" their public image on sites like these, so they're unlikely to discover false accusations against them until they either experience social consequences of those false allegations or they're lucky enough that someone who knows them well discovers it early and sticks up for them.