Concerns about AI-generated disinformation are sometimes downplayed, with the argument that AI only helps create content but does not spread it. However, in 2023, examples showed that AI can automate both creating and spreading disinformation. A developer named NEA PAW created COUNTERCLOUD to test a fully automated disinformation pipeline. In the first step, an AI model scans the internet for articles and selects targets for counter-articles. Another AI model then writes a convincing counter-article, including images and audio summaries. This article is falsely attributed to a fake journalist and posted on the CounterCloud website. Next, another AI generates comments on the article to make it look like there is real engagement. Finally, an AI searches X (formerly Twitter) for related tweets, posts the counter-article in replies, and adds comments. This entire system, which makes disinformation appear authentic, costs about $400 to set up.
Download the medial app to read full posts, comements and news.