It's pretty much what
>>64572 suggested. This spammer is human. She's savvy to how imagebaords work too. She's been targeting imageboards small and large. There are some challenges that make wordfilters and our usual ban evasion deterrent ineffective.
For IP-related automoderation
- The IPs she uses are unique, without any consistent subnets that could be reliably rangebanned
- The IPs are very rarely listed as known VPN or Tor nodes, so the image is able to be uploaded
- Even when a VPN is detected, the spammer simply excludes the image
As far as wording goes
- The words are always obfuscated in unique ways
- The URL shortening service changes almost every day
- The name of the file changes every time (As well as its MD5, and sometimes format and attributes)
And above all, our captcha which is still broken to the point of being even better at deterring spam, is being passed with ease. There are some automated spammers going around but our system catches them for the most part. This one has experience in rapid multi-site spam. She's been using Telegram as a landing site to host the child erotica previews, but despite multiple reports from myself and other board owners, Telegraph hasn't acted towards deleting any of the child pornography they're hosting nor the links leading to the more violent stuff. We're just lucky that the image she's posting here isn't really all that shocking. I've taken the opportunity to confide in our webhost with the details of the spam we're being dealt against our best wishes so it's unlikely they'll pull the rug out from under us in the event that they take a look at our sever during the few minutes that a CP spam is up.