The US presidential elections are at the center of the global information debate and are also attracting the attention of the most up-to-date artificial intelligence systems.
As reported by Repubblica, Elon Musk, one of Donald Trump’s most ardent supporters, has just updated his AI, Grok, capable for the first time of generating images from text.
The first results show how strong the possibility of creating disinformation and manipulation is. A simple feedback to the AI system by writing a text is enough to generate, as in this case, a photo capable of disorienting the American electorate.
“For example, the fake image – Repubblica writes – created with Grok in which a man who looks like Donald Trump and a woman who looks like Kamala Harris are passionately kissing has gone viral. It is one of the many fake – and inappropriate – images shared in the last few hours by users of social X, once known as Twitter, also owned by Musk”.