JOIN USCyber criminals using ChatGPT to create malware, do phishing
With its smart answering capability, OpenAI's ChatGPT (Generative Pre-trained Transformer) bot has taken over the social media platform. People from all walks of life from school-going kids to software engineers are making good use of ChatGPT to write homework essays and even save time writing complex software programmes.
Now, some cybercriminals have taken shine to ChatGPT and are reportedly developing malware to steal financial and personal information on phones and computers.
Though OpenAI has set up security measures to prevent unauthorised use of ChatGPT for illegal purposes, there are loopholes that need to be urgently addressed, says the Check Point Research report.
If a user poses a direct query to ChatGPT to create malware or write a mail impersonating a company for phishing purposes, the OpenAI chatbot won't respond. However. OpenAI in a bid to help companies integrate ChatGPT into their services is offering APIs (Application Programming Interfaces)-- 'text-davinci-003' and the latter lacks any security screening mechanism to prevent misuse.
And, now bad actors are using it to build a Telegram bot that is capable of writing compelling phishing email and even creating a script that can steal PDF documents from an infected computer and sends them to an attacker through FTP (File Transfer Protocol).
ChatGPT-based malicious Telegram bots are apparently on sale on dark websites. Cybercriminals are offering 20 free sample query options and after that, for every 100 queries, people are being charged $5.50.
So far, Open AI has not made any official statement with regard to the misuse of ChatGPT API. People are advised to be cautious when they receive any mail from unknown senders. Never ever share any financial or personal details via email or SMS.
Get the latest news on new launches, gadget reviews, apps, cybersecurity, and more on personal technology only on DH Tech.