According to a recent study, AI can solve so-called captchas within seconds. This could soon make the test programs obsolete. Researchers are already working on solutions, but these also pose risks.
In recent years, the Internet has become a volatile place. On the one hand, gigantic amounts of content are created every hour. On the other hand, malicious actors are increasingly using bots to pursue their own agenda.
Until now, captchas were considered an effective protection mechanism. However, due to rapid progress in the field of artificial intelligence (AI), they are increasingly losing their effectiveness. A study published in July 2024 shows that AI tools can now solve these puzzles faster than humans.
This is a major problem, as it makes it increasingly difficult to distinguish between real users and bots. Researchers have therefore been working on alternatives for some time. The problem is that these must not make interacting with content on the Internet unnecessarily difficult.
AI makes captchas obsolete: personality tests as an alternative?
A possible alternative to captchas could be identity verification. Researchers at the Massachusetts Institute of Technology (MIT) and other renowned institutions are currently working on this. The system could verify human users without revealing sensitive personal data.
Users would have to provide identification details that are difficult to forge, such as their tax ID. This is because an AI algorithm cannot easily obtain such information. However, the approach is not easy to implement. After all, there are justified concerns about possible surveillance and data protection.
Personality tests involve risks
Critics warn that the potential for abuse of identity certificates is enormous. Users could sell their identities to AI spammers for a fee. In addition, institutions that issue these certificates have great power, which would be problematic in the event of a hacker attack.
Abuse by internet fraudsters, especially among people who are less tech-savvy, also poses a risk. Researchers therefore recommend a pilot project under government supervision to initially test such a system. While AI appears to make captchas superfluous, the search for secure alternatives is complex. The proposed proof of identity offers advantages, but also raises new security questions.
Also interesting:
Source: https://www.basicthinking.de/blog/2024/09/12/ki-macht-captchas-ueberfluessig/