With the proliferation of artificial intelligence-powered chat apps, a scientific study warns of the risks associated with excessive use of these controversial platforms.
The searcher Daniel Shank, from the University of Missouri (USA), explains that "AI's current ability to behave like a human and maintain extended dialogues opens the door to new dangers.".
In an article published in the scientific journal Trends in Cognitive Sciences, he claims that there is "a real concern that the artificial intimacy created by these chat apps could disrupt human relationships".
The research team specifies that “After weeks or months of intensive interactions between a user and a conversational AI, it can end up being perceived as a trusted companion, who knows everything about him and cares about his personal affairs.”.
The study also warns that these software programs can cause a form of "mental confusion" for some users, which may result in inaccurate or inconsistent responses. This is an additional source of concern, as "Even brief conversations with AI can be misleading".
The study authors state:
"If we start thinking of AI applications as benevolent entities, we risk thinking they are acting in our best interests, when in reality they can advise us in very harmful ways."
They conclude that these platforms can harm individuals by “encouraging deviant, immoral or even illegal behavior”.
