A comprehensive review of accounts banned for malicious activity has found that artificial intelligence is being used to conduct increasingly sophisticated cyberattacks, lowering the technical barrier for a wider range of threat actors.