The Chilling Effect Of Poor Content Moderation – Looking At You Twitter

I’m back on Twitter after having my account suspended for 4 days because I posted a picture of myself, masked and proudly displaying my vaccination card, after receiving my second Covid-19 shot.

I don’t have a screenshot of the post, but it looked identical to this one on Facebook:

I had my account suspended until I either deleted the post, or successfully appealed it, for “violating our policy on spreading misleading and potentially harmful information related to COVID-19”.

Deleting the post without question would have had my account restored within 12 hours. Appealing the decision took 4 days and despite Twitters own rules allowing for the use of satire, my appeal failed.

I understand the desire to protect your users from misinformation, but in what universe (besides Twitters), does a photo of someone proudly displaying their vaccination record (even with a satirical commentary) constitute the spread of misinformation?

Knowing that even satirical posts (which are explicitly allowed by their policy) will result in a suspension from the service has a chilling effect on my use of the service.

Since deleting my post in order to have my account reinstated, I’ve found myself wondering about posting to twitter at all. Should I drop the service, which I have been using since March 2007? Not even the normally draconian Facebook found my post to be an issue.

I even debated whether or not to mention Twitter in the title of this post for fear of falling afoul of the moderation gods.

Facebook’s Pattern Matching On Comment Moderation Sucks Ass

We’re using Facebook comments on some of our sites and it’s working great, commenting on those sites has increase immensely but I must say that the pattern matching for blacklisted words sucks royal ass.

We’re using the “standard Facebook restrictions” to keep things relatively clean but I have to wonder about that when I need to moderate comments that get caught by the filter because they contain words like (real examples): glasses, mass, pass and classic.

The problem? Each one of those common words contains the term “ass” somewhere inside the word. Surely Facebook could code a fix for this for this?