Cybercriminals have found a way to bypass restrictions set by OpenAI’s ChatGPT to create malicious content such as phishing emails and malware. According to Check Point Research’s latest findings, cybercriminals are exploiting OpenAI’s API by creating Telegram bots that use the API to generate malicious content and steal your data.
ChatGPT has put several restrictions in place to prevent the abuse of the language models. It is important to note that if a user asks ChatGPT to write a phishing email or create malware, it will not generate it due to the restrictions. However, cybercriminals have found a way to bypass these limitations by using the OpenAI API through Telegram bots. These bots are reportedly advertised in hacking forums to increase the engagement.
The cyber threat intelligence company states that the current version of OpenAI´s API has very few anti-abuse measures in place. As a result, it allows malicious content creation, such as phishing emails and malware code, without the limitations or barriers that ChatGPT has set on their user interface.
The CPR in its research found a cybercriminal advertising a Telegram bot using OpenAI’s API without any restrictions and limitations. The company highlights that as part of its business model, cybercriminals can use ChatGPT for 20 free queries and then they are charged $5.50 for every 100 queries.
This serves as a warning to individuals and organizations to be vigilant against phishing attempts and to regularly update their security measures. It also highlights the need for better anti-abuse measures in the development of AI language models.
OpenAI has yet to comment on the findings by Check Point Research and the company’s efforts to prevent the abuse of its API.
ALSO READ l Google Search boss says AI chatbots can give ‘convincing but completely fictitious’ answers
