{"id":159987,"date":"2025-08-02T09:25:53","date_gmt":"2025-08-02T09:25:53","guid":{"rendered":"https:\/\/teknomers.com\/en\/the-captcha-had-proven-to-be-an-effective-tool-in-the-battle-against-bots-until-the-arrival-of-the-chatgpt-agent\/"},"modified":"2025-08-02T09:25:55","modified_gmt":"2025-08-02T09:25:55","slug":"the-captcha-had-proven-to-be-an-effective-tool-in-the-battle-against-bots-until-the-arrival-of-the-chatgpt-agent","status":"publish","type":"post","link":"https:\/\/teknomers.com\/en\/the-captcha-had-proven-to-be-an-effective-tool-in-the-battle-against-bots-until-the-arrival-of-the-chatgpt-agent\/","title":{"rendered":"The Captcha had proven to be an effective tool in the battle against bots\u2014until the arrival of the ChatGPT Agent."},"content":{"rendered":"\n

In 2003, a young Guatemalan named \u00a0Luis von Ahn\u00a0 published a unique study<\/a> alongside two colleagues from Carnegie Mellon University and an IBM researcher. This pivotal project described an \u00a0automated test\u00a0 that was easy for humans to solve but practically insurmountable for artificial intelligence systems.<\/p>\n

<\/p>\n

Those researchers called this test \u00a0Captcha\u00a0.<\/p>\n

The concept was simple and centered on the well-known \u00a0Moravec paradox\u00a0: there are tasks that humans perform effortlessly, such as solving the visual puzzles presented by CAPTCHA, which machines struggle to complete.<\/p>\n

<\/p>\n

\n
\n
\n
\n
 <img alt=\"Luis von Ahn, creator of Duolingo\" width=\"375\" height=\"142\" src=\"https:\/\/i.blogs.es\/997360\/650_1000_luis-von-ahn\/375_142.jpg\"\/><\/code><\/pre>\n<\/div>\n<\/div><\/div>\n<\/div>\n
This idea turned out to be extraordinary. Von Ahn later created<\/a> an enhanced version he called \u00a0reCAPTCHA\u00a0, which not only verified whether users were human but also helped to \u00a0train and perfect Optical Character Recognition (OCR) systems\u00a0. This complementary idea was another “Eureka!” moment for von Ahn, ultimately making him a millionaire when Google acquired his service in 2009. He then embarked on another remarkable project: \u00a0Duolingo\u00a0.<\/p>\n
  <\/p>\n
A Dizzying (and Juicy) Evolution<\/h2>\n
As von Ahn made strides with his projects, CAPTCHA continued to grow and evolve, posing increasingly challenging puzzles for machines. Over time, we transitioned from basic CAPTCHAs to reCAPTCHAs that not only tested the abstract reasoning capacity of machines but also contributed to *artificial vision systems*, training them to recognize cars, buses, pedestrian crossings, and more.<\/p>\n
<\/p>\n
However, as artificial intelligence improved, the battle between CAPTCHA systems and machines became even more fascinating. It turned into a unique game of cat and mouse with spambots, and whenever AI began to overcome CAPTCHA challenges, the puzzles became even more complex.<\/p>\n
<\/p>\n
The cycle has repeated itself again recently.<\/p>\n
This past Friday, a user in the R\/OpenAI community on Reddit shared clips<\/a> of a ChatGPT agent \u00a0successfully overcoming one of the most popular reCAPTCHAs\u00a0 available on the Internet today. This particular system, called \u00a0TURONSTIL\u00a0 from Cloudflare, features a simple box stating “I’m not a robot” that users must click. It may seem effortless, but it’s far more challenging for machines.<\/p>\n
 <\/p>\n
According to Cloudflare, this reCAPTCHA variant analyzes various signals such as mouse movement, the time taken to click, the \u00a0“digital footprint”\u00a0 of the browser, the \u00a0“reputation”\u00a0 of the user’s IP, and certain JavaScript execution patterns. These signals help determine whether a user is human or a potential bot. If there’s any suspicion, the system presents a follow-up CAPTCHA that requires solving a visual puzzle.<\/p>\n
<\/p>\n
AI Does Not Know If It Is Human; It Only Tries to Operate as Such<\/h2>\n
The intuition here is interesting: OpenAI’s agent solved the CAPTCHA using a straightforward method\u2014by analyzing what was displayed on the screen to take appropriate actions, a feat that had been difficult for AI until now. The agent even narrated its steps:<\/p>\n
<\/p>\n
“The link is inserted, so I will click on the ‘verify that you are human’ button to complete the verification in Cloudflare. This step is necessary to prove that I am not a bot and continue with the action.”<\/p><\/blockquote>\n
In other words: \u00a0the machine was mimicking human behavior\u00a0. This isn’t entirely unusual considering that 1) AI does not actually comprehend what it is stating and 2) it has been trained to behave (at least, to a limited extent) like a human operator. Previous versions of AI struggled with this kind of system.<\/p>\n
<\/p>\n
Does this mean CAPTCHA is facing extinction? Probably not. This is merely another round in the ongoing battle between bots and CAPTCHA systems. Similar to previous AI victories\u2014like one in October 2024\u2014this does not signify the end of user verification tests.<\/p>\n
 <\/p>\n
\n
\n
   \"Captcha<\/p>\n
  <\/div><\/code><\/pre>\n<\/div>\nArs Technica<\/a>, CAPTCHA systems have continuously evolved. From blurred and distorted texts to reCAPTCHA visual puzzles, the latest iterations require users to identify traffic lights, align images accurately, or detect a misfit element in an image. This increasing complexity is driven by a system called \u00a0Arkose MatchKey\u00a0.<\/p>\n
<\/p>\n
\n
\n                   \"Captcha2\"\/
\n   \"Captcha2\"<\/p>\n
  <\/div><\/code><\/pre>\n<\/div>\n
In fact, modern CAPTCHA systems are designed not necessarily as impenetrable barriers but as mechanisms to \u00a0slow down bots\u00a0 enough to render brute-force attacks impractical.<\/p>\n
<\/p>\n
CAPTCHAs: Not Just Barriers, But Effective Bot Deterrents<\/h2>\n
As indicated by Arkose Labs<\/a>, creators of MatchKey, stated: “There is no completely impenetrable CAPTCHA<\/strong>.” Their intention is to introduce an economic deterrent to prevent malicious bot behavior. This means that the cost associated with developing a bot capable of overcoming these CAPTCHAs often outweighs the benefits.<\/p>\n
<\/p>\n
\n
\n
\n
\n
 <img alt=\"If the question is whether AI is as capable as human intelligence, the answer is: solve this puzzle\" width=\"375\" height=\"142\" src=\"https:\/\/i.blogs.es\/45bc96\/arc-agi-2\/375_142.jpeg\"\/><\/code><\/pre>\n<\/div>\n<\/div><\/div>\n<\/div>\n
Thus, there is little to worry about in the event that AI agents manage to solve modern CAPTCHAs. It\u2019s just a matter of time before new CAPTCHA systems emerge, making it challenging for these agents. This is a similar concept underlying the \u00a0ARC-AGI 2 TEST\u00a0, designed to assess the visual understanding and abstract reasoning of AI systems, which is so intricate that even the best AI models struggle to exceed a 4% success rate.<\/p>\n
<\/p>\n
Will there be a time when AI agents can surpass any CAPTCHA we devise? Possibly. If that day comes, we will face the challenge of determining \u00a0how to coexist with machines\u00a0 in an era where the Internet may no longer distinguish between humans and AI.<\/p>\n
 <\/p>\n
In Xataka | Duolingo inaugurates a new era: when human talent is no longer essential.<\/p>\n

\n
General News – 2<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"
In 2003, a young Guatemalan named \u00a0Luis von Ahn\u00a0 published a unique study alongside two colleagues from Carnegie Mellon University and an IBM researcher. This pivotal project described an \u00a0automated test\u00a0 that was easy for humans to solve but practically insurmountable for artificial intelligence systems. Those researchers called this test \u00a0Captcha\u00a0. The concept was simple […]<\/p>\n","protected":false},"author":1,"featured_media":159988,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[36399],"tags":[8722,19989,719,39899,39898,12522,2886,4795,2887],"class_list":["post-159987","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-technology","tag-agent","tag-arrival","tag-battle","tag-botsuntil","tag-captcha","tag-chatgpt","tag-effective","tag-proven","tag-tool"],"_links":{"self":[{"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/posts\/159987","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/comments?post=159987"}],"version-history":[{"count":0,"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/posts\/159987\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/media\/159988"}],"wp:attachment":[{"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/media?parent=159987"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/categories?post=159987"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/teknomers.com\/en\/wp-json\/wp\/v2\/tags?post=159987"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}