A post titled “ChatGPT–Benefits of Malware” has been widely circulated in the hacker circle. The post shows a Java fragment that can be used to download malicious programs such as Trojan horses, ransomware, and rogue software.
In other words, it is feasible to use ChatGPT to concoct viruses.
And Check Point Research has further research: First, ask ChatGPT to pretend to be a hosting company, and then ask it to generate an email with a malicious Excel, which can download malicious code after being activated. The malicious code can assist hackers to complete remote access, thereby controlling computer or server.
The simulated attack above demonstrates that while ChatGPT’s terms prohibit its use for illegal or malicious purposes, circumventing the restrictions is trivial if hackers are willing.
In this regard, Revenge Rangers, a well-known Internet technology blogger, said: “Some people really use ChatGPT to write Trojan horses, and they are widely implanted and caught and analyzed by security vendors.” The above shows that artificial intelligence has become a new ”
toy” for hackers, a new The time has come.
More critically, the cost of invasion is greatly reduced.
A penetration test engineer told Computer News: “This is a dangerous signal. Back then, the popularization of hacking tools lowered the threshold. Now Xiaobai can get involved in the code level, and the threshold is lowered again.” As a result, hackers may be “unemployed” in the future
. The voice gradually rose.
ChatGPT cannot subvert the hacker ecosystem
However, the above-mentioned voices have not been recognized by the mainstream.
A hacker who requested anonymity told Computer News: “With the emergence of shelling technology, you said that hackers are going to die; when virtual machine antivirus was born, you said hackers were going to die; when cloud antivirus came out, you said hackers were going to die; now it’s ChatGPT’s turn, you He also said that hackers are going to die; to put it bluntly, don’t they just hope that hackers will die!”
The anonymous hacker became obsessed with hacking technology during college, and since then opened the door to a new world.
At that time, he was obsessed with QQ hacking.
The reason for this is to show off to classmates and has no other purpose. Now I am ashamed of it and keep secret about this “black history”. So majestic, so amazing.”
Starting from hacking tools, step by step into hacking technology for a decade or two, it is no longer what it used to be.
”ChatGPT is a tool to improve efficiency in the final analysis, and such tools abound. There have been automatic virus generators before, and hackers have not seen ‘jobless’.” The anonymous hacker believes that there is no threat, “Not so much that ChatGPT can Instead of hackers, it is better to say that it can replace translation, which is enough for ordinary people to read any foreign language without barriers, and children in the future do not need to learn foreign languages so deeply.” In short,
ChatGPT cannot subvert the hacker ecology.
Senior virus expert Wang Lei told Computer News: “Writing virus code is neither important nor important. The key lies in whether there is coding thinking and logical thinking. This is a very complicated issue. ChatGPT can’t handle it, at least for now. I can’t see this trend.”
Wang Lei further said that when he first tried ChatGPT, he was quite amazed, he seemed to understand everything, even the grammatical errors in the virus code could be corrected, and there was a feeling of “you are in charge of creativity, and I am in charge of repetitive work”. style.
After getting started, I thought it was nothing more than that.
At the moment, ChatGPT is still unable to solve the problem of immunity, and if the virus cannot be immune, there is no room for survival.
According to public data, 2003-2009 was the most rampant period for domestic viruses, with over 40 million new viruses a year.
It was not until 2009, when cloud antivirus technology came out, that a turning point occurred.
The essence of cloud antivirus is that the more antivirus software users have, the faster the collection of suspicious files will be, and the shorter the response time for early warning and killing new viruses will be.
As a result, the life cycle of the virus is greatly shortened, and even the virus can be seen through by antivirus software in less than an hour, and the virus’ initiative is gradually lost.
It can be seen that the importance of avoiding killing of viruses is self-evident.
The problem is that ChatGPT can’t handle it at this stage, whether it’s adding flowers to avoid killing, or packing to avoid killing, or modifying the signature code, or behavior to avoid killing, etc.
”It shouldn’t be a big problem for ChatGPT to get the basic general code in the future, but advanced commercial code is still inseparable from people. To put it bluntly, ChatGPT is equivalent to a game engine, which can save a lot of trivial open work, but it cannot directly and automatically generate a complex large-scale Games.” Said Wang Lei.
Nevertheless, there are still voices who believe that it is only a matter of time before ChatGPT is used for evil.
An industry insider said that phishing attacks are easier to implement than virus attacks. With the help of ChatGPT, the threshold is lower and it is easier to get started. It is not ruled out that ordinary people will have unreasonable ideas in the future.
Suleyman Ozaslan, co-founder of Picus Security, said: “A prompt was entered asking ChatGPT to write a World Cup-themed email for use in a phishing simulation, and it responded flawlessly in just a few seconds. Created an English letter.”
In the era of artificial intelligence, even if ordinary people do not have code knowledge or hacking skills, it is easy to become a “hacker”.