2023-05-02 14:29:47
Artificial intelligence-based chabots are responsible for dozens of spam sites that are beginning to invade the Internet. In a recent report, 49 websites entirely written by AI were identified, which are capable of generating posts, articles and banner ads with the hope of attracting human customers who will fork over money, or even worse, to install spyware and adware. These are real “farms” of spam, generated directly by machines: a risk on which experts had already placed their attention, and which now, with the widespread diffusion of OpenAI’s ChatGPT, are beginning to become a reality. According to NewsGuard, a company that monitors the reliability of news sites, this is just the beginning: if today there are 49 sites discovered, in the very near future there could be thousands.
These sites, which are often impossible to trace back to any human property, produce a huge volume of content related to the most diverse topics, including politics, health, entertainment, finance and technology, obviously based on fake news or with a narrative steered in a certain direction. The articles can be recognized thanks to their elementary grammar and a certain repetitiveness, trademarks of ChatGPT writing, but at the same time there is the risk of mistaking them for real articles. For example, in early April, an AI-generated site, CelebritiesDeaths.com, published the news of Joe Biden’s death. On these sites there is a huge amount of banner ads and links that lead to purchases.
#designs #spam #sites #ChatGPT