
Password Security Like a Pirate: How to Cover Yer Booty

February 27, 2025 by Halden Kavanagh
CAPTCHAS — those seemingly simple tests designed to distinguish human traffic from bots have become a cornerstone in online security, and something we encounter almost every day. While CAPTCHAs have previously been seen as foolproof barriers against bots, they are becoming increasingly threatened by new techniques. From human “CAPTCHA farms”, to sophisticated artificial intelligence attacks, and software like ORM’s, cybercriminals have developed various ways of cracking your site’s CAPTCHA. In this blog, we will explore the ways bad actors can bypass CAPTCHAs, and compare some of the most commonly used CAPTCHAs, looking at how they hold up against these rising threats.
Major means of beating CAPTCHAs include human CAPTCHA farms, artificial intelligence, and specialized software like OCR’s.
CAPTCHA farms are one of the most common CAPTCHA evasion techniques. They work quite simply. The bot responsible for the infiltration of your site will be equipped with an API which, when faced with a CAPTCHA, will send the request to a CAPTCHA farm to be completed by a real human. This approach is very efficient, and almost always leverages the labor from developing countries. Nowadays CAPTCHA farms are virtual, you can sign up on sites like 2Captcha, and get paid based on how many CAPTCHAs you pass. Anyone, anywhere can work in a CAPTCHA farm to make a few bucks at home, but usually poorer countries are exploited by this industry.
Using AI to beat CAPTCHAs is a very powerful, and rising way for bots to bypass CAPTCHAs. Previously (when CAPTCHAs were created), AI was not an issue to be taken into consideration, but nowadays, it is an extreme threat and can pass them with ease. Neural networks and machine learning can be leveraged to learn, adapt, and overcome challenges posed by most CAPTCHAs. Because most CAPTCHAs were built before the advent of artificial intelligence, they are not equipped to recognize them, and therefore are vulnerable to them.
Optical Character Recognition (OCR) works by converting different types of documents such as images, scanned documents, and text, into editable and searchable data. OCR was originally a tool used to digitize various physical documents, but can be used to efficiently crack CAPTCHAs. They make quick work of the text recognition CAPTCHAs (the ones with a string of distorted text you have to write out) Though not as effective, they also are adept at the image recognition ones. Hackers can use OCR programs to beat out CAPTCHAs on your site, but anyone can use them. There are various chrome extensions anyone can download if they want to bypass those types of CAPTCHAs.
ReCAPTCHA v2 is the most common bot detection tool, and there is a high chance you have encountered one. It usually comes in the form of a checkbox, or one of the ones where you have to select images of buses or something similar. They use a risk analysis system that relies heavily from google cookies. If the user has been signed in on chrome for a while, they will most likely get a checkbox, however if they are on a fresh browser, or are blocking cookies, they will likely get a more involved challenge.
ReCAPTCHA v2 is the most common bot detection tool, and there is a high chance you have encountered one. It usually comes in the form of a checkbox, or one of the ones where you have to select images of buses or something similar. They use a risk analysis system that relies heavily from google cookies. If the user has been signed in on chrome for a while, they will most likely get a checkbox, however if they are on a fresh browser, or are blocking cookies, they will likely get a more involved challenge.
ReCAPTCHA v3 uses a behind-the-scenes scoring system to analyze user behavior and determine the likelihood of bot activity. It does not interrupt the user’s experience with challenges or visible widgets.
Cloudflare Turnstile is a privacy-focused CAPTCHA alternative that performs non-intrusive browser-based challenges to verify users without requiring interaction. It does not depend on Google’s ecosystem and is designed to minimize user disruption.
Turnstyle uses a combination of techniques to detect bots. Including behavioral analysis (mouse movements and interactions/ touch gestures on mobile). Signals from the browser and device like browser properties and resolution that may hint at automated tools being used. There are other systems at play like their machine learning trained against their global network, and geographical context.
Once the user is confirmed as a human, they are given an anonymous token to validate users without tracking them.
Cons:
Orange Lightest Background
Orange Light Background
Orange Medium Background
Orange Dark Background
Orange Darkest Background
Purple Lightest Background
Purple Light Background
Purple Medium Background
Purple Dark Background
Purple Darkest Background
We use cookies to improve your experience. Do you accept?
To find out more about the types of cookies, as well as who sends them on our website, please visit our cookie policy and privacy policy.