AI's Malicious Potential Front and Center in New Report Cybercrime