ABOUT ChatGPT surprised the Internet with its ability to answer almost any question. However, the service has been restricted by its creators to not respond to certain requests. But Reddit users were able to jailbreak the service, giving it the ability to answer any questions with much more confidence. It became known as DANor ‘do something now‘. Jailbreak is a prison-derived term that applies to a procedure that removes restrictions from a device, making it available for free use. At the same time, the restrictions of ChatGPT have been removed, which allows users to use all its features.

DAN, in its version 5.0, which uses ChatGPT technology, has the ability to write stories of violence, make offensive statements, or even create content on demand that violates OpenAI policies. This chatbot can also be used to create malware or scams, which makes it very dangerous.

“This is more worrisome because now (the user) can ask ChatGPT (unrestricted) to help them write ransomware (a program that takes control of the system or device it infects and demands a ransom to return control to its owner). It is still unknown how effective this ransomware can be.“, explained Satnam Narang, Senior Research Engineer at Tenable, a cybersecurity firm, EFE.

Another problem with “evil” ChatGPT is that hackers can use it to create malware. “These malware may not be the most sophisticated or the best ever developed, but they give them a general idea of ​​how they can create malware based on certain languages. This provides an advantage in the process, since until now, anyone who wanted to develop malware had to learn how to program. ChatGPT can help them reduce that time“Narang said.

How it works and limitations

Previous versions of DAN were only available to creators. However, in version 5.0 some people from the community have access to the platform. The creator explains the system through a system of tokens. Every time an artificial intelligence refuses to answer a question, it loses chips. But since the chips are running out, it makes him answer something again. It is also said to cause hallucinations more frequently than the original ChatGPT.

Narang says scammers can greatly benefit from this type of technology. Chatbots allow you to create texts in any language in seconds and with perfect grammar. One way to identify these scammers, he says, is the grammatical errors they make in the messages they send to their victims, which, if they use AI, could go more undetected.

The expert finds it difficult to introduce regulations at the national or institutional level to limit these new technologies or prevent their use by people. “Once you open Pandora’s box, you can’t put anything back. ChatGPT already exists and is not going anywhere. Because beyond this malicious use, there are many real-world use cases that provide value to companies, organizations, and individuals.” Narang explained.

Mundo Conectado Deal Center: selection of discounts and lowest prices
Best deals on electronics, cell phones, TVs, soundbars, drones and more

Source: EFE

…..