Malicious code related to ChatGPT blooms on Facebook

## Meta said that since March, it has detected and blocked more than 1,000 malicious links that are advertised as having ChatGPT integration. According to Facebook’s parent company, hackers create a series of malicious software in the name of tools that use ChatGPT to attract interest and lure users to download. From there, bad guys can gain access to the device, steal important information and user money. Since March, Meta has detected about 10 different types of malicious code and blocked more than 1,000 malicious links advertised as AI-powered chatbot integration tools. Guy Rosen, Meta Information Security Manager, says the company is getting ready to deal with new AI abuses like ChatGPT. Previously, some US lawmakers said that bad guys can take advantage of AI chatbots like ChatGPT to manipulate and transmit fake and distorted information to serve a certain campaign, such as elections. According to Rosen, it is still early to evaluate the use of AI in propaganda activities, but sooner or later, the bad guys will also use this technology to accelerate and scale to achieve their goals. (ChatGPT)(https://vnexpress.net/chu-de/chatgpt-5974) was introduced by OpenAI in November 2022 and quickly caused a fever with more than 100 million users. This AI works based on the NLP (Natural Language Processing) model, trained with large data sources on the Internet to answer users’ questions. Although well-received, ChatGPT also causes concern when it often gives wrong answers, or is at risk of being used for malicious purposes.