CAPTCHA tests (short for Completely Automated Public Turing test to tell Computers and Humans Apart) were originally introduced to the internet in 2001 by Yahoo! and created by Luis von Ahn, according to the National Inventors Hall of Fame. Ahn also helped create the reCAPTCHA test, where users must type in the word or letters they see.
But now it seems that those CAPTCHA tests are getting harder and harder for us to solve.
Why were CAPTCHA tests made?
According to Google, CAPTCHAs are a security feature that helps protect your accounts from entities that are trying to break in to your account and decrypt your password. Multiple companies use them to prevent bot attacks that can crash their websites and create security risks, per The Wall Street Journal.
Captcha tests originally worked by having users type in letters or numbers they may see in a picture, according to Cloudflare. Bots are now able to decipher these simple tests, so more complex captchas have been created, including image recognition tests, checkboxes and user behavior assessments.
How they’re getting harder
Business Insider says the “robots are making it harder to prove you’re not a robot.” Multiple people have shared their experiences with CAPTCHAs online, leaving some to wonder why they’re getting so hard to pass.
According to The Wall Street Journal, current bots have become better at labeling images, so we can’t just rely on image-identification CAPTCHAs.
Arkose Lab’s CEO, Kevin Gosschalk, told The Wall Street Journal, “So now enters a new era of Captcha — logic based.” So instead of just identifying certain images, users will have to do something with the test information, such as rotating or moving an object.
Some new CAPTCHAs are including AI images, which, according to The Wall Street Journal, are more difficult for robots to identify but are just baffling to people. Gosschalk explains that all CAPTCHAs will soon be solved by a robot at some point, but what CAPTCHA developers are really going after is “to design something that’s really expensive for developers to try and train software to do” with new CAPTCHAs.
Per The Atlantic, AI and bot machines can solve CAPTCHAs better than most people since becoming smarter than before. With original CAPTCHAs that required us to just type in words or letters, bots can now parse text easily.
Then CAPTCHAs shifted to image identification, but the bots were soon able to recognize those, too. Finally, CAPTCHAs have reached the point of using weird photos and tasks to differentiate human users and bot users.