Over time, technologies such as ChatGPT will be adapted to various production processes, thanks to their openness to plugins in third-party environments that will fuel its power for data management and content creation. This, like all innovation in the world techhas advantages as well as risks. A recent report by the European Union Agency for Police Cooperation (Europol), sheds light on a latent danger in the application of this generative AI: the phishing.
In the document “ChatGPT: the impact of broad language models on law enforcement’, the authority mentions that there is a risk of vulnerability with the use of a system that, by accessing a huge database at the request of a user, can create more refined criminal solutions to carry out scams.
The potential of this tool for crime lies in its ability to generate content at the request of each user, creating multimedia content and text that can convince people on social networks or messaging services. The first virtue is the possibility for criminals to improve spelling, a detail not minor to identify false conditions in scams through networks and apps.
Fake News Customized: the risk of ChatGPT
The other risk evident in the massification of ChatGPT is the ability of this new technology for the creation of images, audios and videos related to artist-endorsed offers and promotions or, in the worst case, the use of the deepfake for blackmail.
In general, scams involving images of high-impact personalities on social networks tend to have a larger number of victims. From messages shared using fake profiles and varying the name of the official profile by one or two letters, to cloning bank websites and streaming services, criminals take advantage of these opportunities to extract credit cards, passwords and confidential information from victims.
Finally, Europol warns that OpenAI, the developer of GPT-4, has not taken the necessary measures to prevent the use of its powerful tool in criminal activities. For this, the European entity’s Innovation Lab to train different law enforcement agencies on new ways to use GPT in advanced techniques using generative AI.
