Cybercriminals can’t agree on GPTs

Cybersecurity firm Sophos has detailed how cybercriminals are turning large language models (LLMs) to their sinister uses. The firm has reported the technology has been used to generate, among other things, email lures, dynamic phishing landing pages and even malware. Sophos cited WormGPT and FraudGPT, two LLM services, as reading like “an AI promotional piece”.