Meta's security teams have uncovered 10 malware families using ChatGPT-themed lures to spread malware on its platforms.