Top News

AI Threats: Artificial Intelligence was created to help humans, but now this technology is also creating new concerns. Recently, in a test, Claude 3 Opus discovered the weakness of the browser and devised a way to hack it without any human help. This incident has sparked a new debate in the tech world regarding the safety and ethics of AI.

How AI discovered Chrome’s weakness?

In this experiment, a cyber security researcher gave the task to AI to find a flaw in the V8 engine of Google Chrome. Usually, it takes weeks or months for expert hackers to do this kind of work. But the AI ​​gradually learned and not only found the bug, but also took advantage of it and prepared a complete “hacking plan” i.e. exploit.

Failed in the beginning, then did wonders

In the beginning the AI ​​was repeatedly making mistakes and the code was failing. But he learned from his own mistakes and improved the code and eventually created a working exploit. This shows that AI can not only learn, but also solve complex problems while improving itself.

What exactly is an exploit?

If understood in simple language, exploit is a method by which a software’s weakness is taken advantage of to infiltrate the system or gain control over it. In this case, the AI ​​targeted the part of Chrome where the probability of error was higher.

Expensive but dangerous sign

This entire test cost approximately $2,283 i.e. lakhs of rupees and billions of tokens were used. This method is certainly expensive at the moment, but it clearly indicates that if AI becomes cheaper in the future, it may become easier to misuse it.

Old software can become a big threat

The special thing about this attack was that it was done on the old version of Chrome. Even today, many apps use old browser engines, which are not updated. This is why such systems can become easy targets for AI.

Why are experts worried?

This incident has worried tech experts. If the power of AI continues to increase like this, then tasks like hacking may become even easier in the future. For this reason, many big companies are avoiding making their advanced AI models completely public, so that they cannot be misused.

It is important to be cautious

AI is a powerful technology, but only its proper use can guarantee a secure future. Users should also always keep their devices and software updated, so that any danger can be avoided.