During the initial hype surrounding generative AIs like ChatGPT, it was observed that they were inadvertently generating keys for obsolete Windows versions, such as Windows 7. It seems they're not very good at controlling what they say. But this is exacerbated by recent reports that they are being used to generate ransomware.
The security firm Barracuda has discovered that ransomware attacks doubled between August 2022 and July 2023. This increase was driven by AI-powered phishing campaigns and the automation of attacks to expand their reach, also using AI. They state that, " despite the continued success of traditional attack methods throughout 2023 and beyond, attackers will look to generative AI to develop increasingly effective attacks ."
What do AI designed ransomware attack?
Barracuda believes that the infrastructure, municipal systems, education, and healthcare sectors are considered easy targets. These sectors have virtually limited resources, making them susceptible to paying to protect themselves from attacks. Certain sectors are legally obligated in their countries to report cybersecurity incidents , which makes them more visible in the data.
Ransomware attacks on financial institutions fell from 6% to 1%, due to possible security improvements. On the other hand, software companies were the most popular target and experienced a significant increase in ransomware attacks year on year. These would be an easy target because an attack disrupts the supply chain for these and many other companies.
Fleming Shi, CTO of Barracuda, commented that recent advances in generative AI will help ransomware gangs increase their attack rate with more effective cyber weapons. Given this, they suggest the usual advice: don't skimp on security systems. Because ransomware can now adapt quickly using generative AI , unlike hackers who can modify the code as soon as a way to circumvent it is detected.
Source: TechRadar
