AI is becoming more dangerous - technologies are already being used against people

The article warns about the growing threats associated with artificial intelligence: scammers are already using deepfakes and voice cloning. The author calls for distinguishing real risks from speculation, noting that there is no convincing evidence of mass smartphone eavesdropping. A particular danger is the transfer of AI control in the military sphere.

An article on the Cursorinfo portal, citing Channel 14, warns about the growing threats associated with artificial intelligence. It notes that scammers are already actively using realistic deepfakes, voice cloning, and creating convincing images to deceive victims. The author, Tomer Adoni, calls for distinguishing real risks from frightening speculation. In particular, he notes that there is no convincing evidence of mass smartphone eavesdropping on user conversations for advertising purposes. It is also emphasized that background computing on devices, including federated learning, does not necessarily mean the collection of personal data. Particular concern is raised by the transfer of AI control in security systems and the military sphere, where the final decision should remain with humans.