Bots Outwitting Human Verification Tech


When you spend your hard-earned ad dollars to get your message in front of eyeballs, you are making the assumption that those eyeballs are attached to humans. As it turns out, this is not a safe assumption at all. A click on a webpage does not, in fact, represent human engagement with the ad. If that is how you measure the value of your ad spend, then your number one potential customer is a robot. 

WIRED reports that bot traffic can run as high as 80 percent. That means the biggest challenge Internet advertisers face is making sure they get their message in front of a human audience, and secure human engagement with their ads. That’s why we have the dumpster fire of a disaster known as CAPTCHAs.


A CAPTCHA is a small graphic presenting a collection of letters and/or numbers that have to be precisely entered into a textbox positioned just below the graphic. There was a time when these characters were more like human-readable words. But that was too easy. If humans could easily read it, so could bots. 

In an attempt to defeat the bots, they started scrambling the letters. Still too easy, they decided to distort the characters. They would alter their shapes, run lines through them, and mix them in with random graphics that also looked like letters and numbers. Not only could a human not tell what the characters were, they also couldn’t tell which images were the characters they were meant to decipher. 

That also exacerbated the problem of accessibility. Blind and visually impaired people have been using computers for ages with the aid of accessibility software. But the software was no match for the CAPTCHA. At some point, they introduced an “accessible” CAPTCHA that used sound to speak the characters that were to be typed. 

Unfortunately, bots can also translate spoken characters. So to combat that, they used a system that speaks the characters through a demonic-like noise matrix where it is as difficult to pick out the characters aurally as it is visually. Today, CAPTCHAs are so aggressively hostile that they do the exact opposite of what they were intended. When a human encounters one, they are highly likely to avoid the site altogether, leaving it to the bots: the only things in the universe that can read CAPTCHAs.


The good news is there are some attempts to reinvent the CAPTCHA. One such effort is from RevTap, who created a system they claim will deliver a 100% human audience, no bots allowed. 

They also claim to be able to increase engagement with ads. What they have done is represent the ad in images and logos. Rather than have a person enter an indecipherable code, they have them click on the Coke bottle. In this way, they gamify the verification system. 

Such a system placed at points of contact such as logins and checkouts would definitely increase interaction with the ad, while at the same time ensure a human is on the other side of the interaction. Get more info on the new technology involved in monetizing CAPTCHA ads at


Scientific American addresses eight alternatives to CAPTCHAs, including the image puzzles favored by RevTap. Some of those alternatives include:

  • The math puzzle
  • The trivia puzzle, and
  • Text message verification

All of these have their own set of issues. My two favorites are the timing trick, and the hidden-field scam. The timing trick takes advantage of the fact that it takes a human a certain amount of time to enter text into a field. A bot will do it much faster than a human. The hidden-field scam takes advantage of the fact that bots can read code that humans can’t. If a hidden field, like a phone number, is hidden to users but is filled in anyway, it was done by a bot.

This is a winnable battle. What we are fighting against is our own worst impulses. It is human vs. human, not human vs. tech. It is humans who make the bots. It is humans that make the systems that incentivize the making of bots. When we want to communicate with other humans bad enough, we can usually find a way.