The war against cybercriminals is as old as computer science itself. The techniques used by hackers evolve with technology. With the introduction of ChatGPT, hackers like these Chinese, Iranian and Russian hacking groups are using this cutting-edge technology to orchestrate cyberattacks.
Microsoft and OpenAI thwart five hacker groups exploiting ChatGPT
Microsoft Threat Intelligence and OpenAI have intercepted several hacker groups attempting to exploit ChatGPT-4 Orchestrate cyberattacks. In particular, the hackers used AI to search for targets and obtain information (including their defense systems) about the companies being attacked.
According to OpenAI, the intercepted ChatGPT accounts also used the AI tool to attempt to launch phishing campaigns, translate technical documents, develop social engineering techniques, find coding errors, and more. OpenAI highlights an application that is only at an early stage.
By analyzing the intercepted accounts, experts identified five hacker groups linked to the nation-states. Charcoal Typhoon and Salmon Typhoon would be associated with that China. Crimson Sandstorm would be sponsored byIranEmerald Sleet on North Korea and forest storm Russia.
OpenAI is having trouble with the misuse of ChatGPT
After the launch of ChatGPT, the digital world watched almost helplessly as a tsunami of scams, DeepFake and other AI-generated cyberattacks took place. Still, OpenAI has invested heavily in cybersecurity.
We disrupted the use of our platform by five state-affiliated malicious cyber actors.
Work in collaboration with the Microsoft Threat Intelligence Center. https://t.co/xpEeQDYjrQ
— OpenAI (@OpenAI) February 14, 2024
Among other things, the AI company's task is to implement measures Prevent ChatGPT from providing malicious answers to users, racist or dangerous. Nevertheless, hackers have found ways to do this Breaking out of prison the tool and let it do everything we ask of it.
Given this, OpenAI emphasized that this is necessary Stay one step ahead of evolving threats. Without giving too many details about the approach it will take, the AI company said it will focus on transparency and collaboration with other AI developers, among other things, to secure its artificial intelligence models.
Our blog is reader-run. When you purchase through links on our site, we may earn an affiliate commission.