TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Ask HN: Anyone here working at Meta on bot detection?

2 pointsby pennyintheslotabout 2 years ago
I just want to give you an example of the unbelievable false positives your system creates and how this ultimately hurts your bottom line.<p>1. Yesterday I tried to create an Instagram page for my food-related project to advertise it through social ads. I used my MacBook, my home IP within the EU and a Google workspace email address. The domain was recently acquired and has no negative history. The account name and domain are food-related and completely safe for work.<p>2. After entering the email confirmation code for sign up the account was immediately banned (Error message: “Something has gone wrong”). Ehhh weird. I thought it was a mistake and tried to login with the credentials I had just set.<p>3. The login worked and I was then asked to complete a reCAPTCHA and provide a mobile phone number. OK fine. I completed the reCAPTCHA, provided my number and entered the confirmation code. Then I was asked to upload a picture of my face and a piece of paper with the account name and another code written on it. (?!?!)<p>4. I also complied with this absurd inquiry and uploaded a picture of my face and a piece of paper with the account name and code written on it. 24 hours later I received an email that stated I would not be able to use this account because I didn’t upload an acceptable image. (The image was completely fine and included exactly the details Instagram had asked for!)<p>5. The end result: The account was permanently banned for doing… nothing. There is no way to appeal the decision. The account name that matches my domain is now taken by an account that doesn&#x27;t even exist.<p>All in all this was the most bizarre interaction I’ve ever had with an online platform. Your bot detection seems to be totally off and produce absurd false positives. Why would you instantly ban an account on sign-up and then make the ban permanent after all the human verification requests have been completed dutifully?<p>I used the same email address and IP to create an account on Reddit, Pinterest, Twitter and TikTok without any issues afterwards. Permanently banning faithful actors on sign-up who are looking to buy ads might hurt your bottom line.

no comments

no comments