Post
For the third time this #TrustCon I'm hearing the phrase "trust and safety is adversarial". There is no end state where it's "done". It is for this reason I strongly believe every platform needs to allow an admin to block harmful content, starting with text strings and URLs known to be used in spam and fraud campaigns.
Yes, it's not perfect.
Yes, it may be misused or abused by individual admins.
Yes, it's embarrassing we can't as a network share a signal and protect our hosted users.
Step 1: known malicious URLs. If a message contains a disallowed URL a. Disable the link and strike through b. Flag the message
Step 2. Exact match text. Everyone knows the Scunthorpe problem. We've known about it for decades. This is a basic first step.
Step 3. Known bad media. Admin can flag media, same filename, same size? Disallow it.
Step 4. Understand that there will always be new attack vectors. Iterate and add more common, proven ways to combat harmful content.
A space for Bonfire maintainers and contributors to communicate
