Cybersecurity Alert: Major ChatGPT Account Leak Discovered on Dark Web
Today, cybersecurity firm Group-IB revealed a substantial leak involving ChatGPT account information on the dark web. The Singapore-based company identified 101,134 devices infected with malware, all containing login credentials for the generative AI platform.
Group-IB traced the origins of these compromised devices, determining the regions most affected. The data indicates that the Asia-Pacific region was particularly hard hit, with over 40,999 accounts breached. The Middle East and Africa followed closely with nearly 25,925 affected accounts, while Europe reported 16,951 breaches. Notably, India recorded the highest number of compromised accounts at over 12,632, followed by Pakistan (9,217), Brazil (6,531), Vietnam (4,771), and Egypt (4,588).
Experts at Group-IB pointed to the increasing trend of employees using chatbots for tasks ranging from software development to business communications. ChatGPT by default stores user query and response history, which means unauthorized access to accounts could expose sensitive information, rendering organizations vulnerable to targeted attacks.
To mitigate risks associated with compromised ChatGPT accounts, Group-IB recommends that users regularly update their passwords and enable two-factor authentication (2FA). This additional layer of security requires users to enter a verification code, typically sent to their mobile devices, before accessing their accounts.