Stolen OpenAI credentials prevalent on the dark web

Over 200,000 OpenAI credentials are being sold as stealer logs on the dark web, reports BleepingComputer. Moreover, dark web mentions of OpenAI's artificial intelligence chatbot ChatGPT have exceeded 27,000 during the past six months, indicating threat actors' growing interest in leveraging generative AI tools in cyberattacks, according to a report from Flare. The findings come after a Group-IB report last month showed that more than 100,000 ChatGPT user accounts stolen by information-stealing malware have been traded on the dark web. Meanwhile, the newly developed generative AI tool WormGPT based on the GPT-J open-source large language model was also recently reported by SlashNext to have facilitated the creation of highly convincing phishing messages that could bring upon more effective business email compromise attacks. Organizations have been urged to bolster their defenses against generative AI threats by strengthening employee security training programs and bolstering their processes for email verification.