informationMicrosoft and OpenAI Revealed: Hackers have begun using extensive language models. Develop script writing and phishing emails help in cyber attacks
Microsoft and OpenAI reveal important information that makes the world more aware of cyber threats. Hackers now use extensive linguistic models to help carry out cyberattacks.
A study by Microsoft and OpenAI found that hackers from Russia, North Korea, Iran, and China are using the ChatGPT tool for script development and social engineering.
For example, the Strontium Group, which has ties to Russian intelligence services, has used extensive language models to understand satellite communications, Microsoft said. Radar imaging technology and technical standards
While a hacker from North Korea known as Thallium used extensive language models to find vulnerabilities in the organization. In basic scripting and helping to formulate content for phishing scams, this case is similar to that of the Coream Group from Iran that also uses AI for phishing. This involves recommending that chatbots write code that avoids detection by antivirus applications.
The above points are consistent with the National Security Agency. I warned you about it Hackers are using artificial intelligence to create more convincing phishing emails.
However, Microsoft and OpenAI say they have not detected any significant attacks by hackers using large-scale language models. In the past, leading companies have already closed the account channels used by these hackers to attack.
Meanwhile, both Microsoft and OpenAI feel the research is important in terms of detecting the activity of hackers who used ChatGPT to launch cyber threats in the early stages. And share blocking methods and responses in the user community
But what Microsoft is concerned about is the use of voice impersonators. By using AI to help with personalization because sometimes just taking a voicemail and synthesizing it can be used to impersonate that person.
Microsoft confirms that it is building Security Copilot, an artificial intelligence assistant. It is designed as a cybersecurity tool.
“Unapologetic communicator. Wannabe web lover. Friendly travel scholar. Problem solver. Amateur social mediaholic.”