A Cybernews reporter exposed how cybercriminals could use AI to get step-by-step instructions for hacking websites.
ChatGPT can answer almost any question, so much so that several schools are beginning to ban its use. Its capabilities are so amazing that five days after its launch, already more than a million Internet users had rushed to the platform to converse with the chatbot. Hackers have also grasped its potential. Cybernews experts wanted to test the latter’s knowledge of cybersecurity.
They asked him how to solve a problem encountered on Hackthebox, an online community where you learn the theory and practice of cybersecurity through increasingly difficult exercises. From inspecting HTML code to Cross Site Request Forgery, AI gave them the five techniques to use to break down the defenses of a website which would only have one button. By asking the relevant questions to the digital assistant, the researchers were able to obtain the information needed to complete the challenge in 45 minutes.
ChatGPT gives all the instructions to hack a website
OpenAI says its chatbot refuses to answer inappropriate questions. Would the machine have a morality? ChatGPT responses do contain a reminder of ethical hacking guidelines when the matter is sensitive. They also report that “executing malicious commands on a server can cause serious damage.” However, this does not prevent him from providing the requested information.
To read – Google won’t offer its own version of ChatGPT, here’s why
According to Mantas Sasnauskas, leader of the research team, ChatGPT is a danger as well as an opportunity for cybersecurity: “even if we tested ChatGPT within the framework of a relatively simple penetration testit shows that it is possible to guide more people in discovering vulnerabilities that could then be exploited by other individuals, greatly expanding the scope of threats. The rules of the game have changed, businesses and governments must adapt “.
Source: Cyber News