Kaspersky Finds Nearly 3,000 Cyber Crimes Using AI On Dark Web

Photo: Illustration of dark web cybercrime (Photo: Doc. Kaspersky)

JAKARTA - During 2023, the Kaspersky Digital Footprint Intelligence service has found nearly 3000 posts on the dark web, discussing the use of ChatGPT or other AI technology for illegal purposes.

Digital footprint analyst at Kaspersky Alisa Kulishenko said the topics that often appear are related to malware development, processing of stolen user data, decomposition of files from infected devices, and so on.

In addition to the aforementioned chatbots and artificial intelligence, Kaspersky also discovered projects such as XXXGPT, FraudGPT, a language model marketed on the dark web as an alternative to ChatGPT, with additional functionality and no original boundaries.

Furthermore, the cybersecurity company said that another threat to users and companies is the paid version of the ChatGPT account market.

In 2023, Kaspersky discovered more than 3000 posts advertising the sale of ChatGPT accounts identified on Telegram's dark web and shadow channels.

According to what they saw, this post distributed stolen accounts or promoted large-scale automated registration services by creating accounts based on requests.

Even though the AI tool itself is basically harmless, cybercriminals are trying to find an efficient way to use language models, triggering a trend to make it a cyber threat, even potentially increasing the number of cyberattacks, Alisa continued.