This Is How Those ‘I’m Not a Bot’ Boxes Work

Hint: It has to do with your mouse.

I'm not a robot box check.ninefotostudio/Shutterstock

Back in the heyday of the Internet, people had to verify their non-robot status by typing some squiggly letters onto the keyboard. Now, all you have to do is click a box saying “I’m not a robot.” It might seem like one click of the mouse is easier than typing out letters for robots, but that’s not the case.

Victor Congionti, CEO of Proven Data, explains that the “I’m not a bot” boxes work by identifying user patterns and behaviors to make sure you’re trying to log into your account—and not a bot. “These boxes work by detecting movements in the mouse direction, which only a human would create when using the computer,” Congionti says. People tend to move the mouse all over the place, but bots often make perfect geometric shapes and can’t always mirror human movements. These, however, are the 12 red flags someone may be spying on your computer.

The box also looks at your activity online, whether or not you clicked a lot of these boxes recently or not, and a lot of other little data points to detect whether it is a real person checking that box or some automated bot somewhere, according to Fabian Wosar, CRO at Emsisoft. We suggest keeping an eye out for these 16 other clear signs you’re about to be hacked.

Nowadays, the boxes will also often open up to an image puzzle as well. They might ask you to click on traffic lights, bridges, or cars among a few different images. There are two reasons for this, according to Jamie Cambell, a cybersecurity expert and founder of First, bots have trouble id’ing blurry images and second, in the case of Google, they use real images to help build better Google maps and other products. James Goepel, an adjunct professor of cybersecurity law and strategy at Drexel University, explains that these bot boxes are training artificial intelligence (AI) engines. So as you prove you’re not a bot, you’re helping train or teach technology the difference between bots and people, too.

Although there are all these different security measures to protect your information, bots still try and can click on these boxes or images even though they often fail. To defeat the boxes, criminals have to train a variety of AI engines to look for a specific item like a sign or car. Where they currently have trouble is that a single AI can’t efficiently look for a variety of things, making it harder for them to defeat these security measures. Now that you know everything about bot boxes, check out these 20 other cybersecurity secrets hackers don’t want you to know.

Emily DiNuzzo
Emily DiNuzzo is an associate editor at The Healthy and a former assistant staff writer at Reader's Digest. Her work has appeared online at the Food Network and Well + Good and in print at Westchester Magazine, and more. When she's not writing about food and health with a cuppa by her side, you can find her lifting heavy things at the gym, listening to murder mystery podcasts, and liking one too many astrology memes.