Russian hackers want to steal your ChatGPT account

Chatbot accounts powered by generative AI, such as ChatGPT, are increasingly sought after by cybercriminals. Hackers are working to exploit stolen credentials to gain fraudulent access to accounts for personal gain or to circumvent sanctions.

AI-powered chatbots are a hot commodity. In a report Published on September 18 by cloud company Sysdig, it is reported that accounts with access to language models are sought by cybercriminals. Tools that the general public is familiar with: ChatGPT is based on these technologies.

According to the group’s cybersecurity experts, hackers are particularly looking to trap employees to recover their personal or professional accounts. One of the reasons is linked to the geopolitical context, and more specifically to the wave of access restrictions that has hit Russia since the extension of the war in Ukraine in 2022.

Russian hackers want to steal your ChatGPT account
An example of an email that could be used as phishing by a hacker. // Source: Numerama

Indeed, numerous sanctions have been imposed by the West, which makes access to tools and technologies much more complicated for Russia. Technology companies such as Amazon and Microsoft have banned all organizations based in Russia from using their tools. A phenomenon that concerns

Accounts hacked just to chat with ChatGPT, too

Since then, hackers have been trying to access various chatbots, or conversational agents in French, such as ChatGPT, Claude or NovelAi. This is in order to take advantage of the most recent advances in generative AI on these platforms. OpenAI has also reported the hijacking of its tools by cybercriminals.

« We have observed many other examples of Russian queries ” note cybersecurity experts. Russian citizens are also reportedly trying to illegally access corporate accounts to use chatbots as well.

In particular, the team observed a Russian student using stolen credentials from Amazon Web Services to create a model of Claude, a chatbot developed by the American company Anthropic. The student was working on a project involving AI-based chatbots. Sysdig experts believe he used stolen credentials and an external bot to avoid being traced.

Researchers have noted that some hackers have sometimes much lower ambitions. Thus, it has been noted that some attempts to gain access were just aimed at accessing these conversational agents to dialogue with them, and sometimes to have erotic interactions. Loneliness can be a scourge among hackers.

Follow the latest news on artificial intelligence with our newsletter!


Source: www.numerama.com