Skilful voice impersonators are able to fool state-of-the-art speaker recognition systems, as these systems generally aren’t efficient in recognising voice modifications, according to new research from the University of Eastern Finland. The vulnerability of speaker recognition systems poses significant security concerns.