Analysis
5 Ways cybercriminals are using AI: Malware generation
Publication date not yet evaluated · Added · 3 reports barracuda.com
Page published
Coverage timeline
Why it matters
Barracuda's overview helps defenders understand how threat actors are operationalizing generative AI for malware creation and abusing LLMs, framing the emerging attack surface around dark LLMs and model poisoning.
Barracuda's blog series examines how cybercriminals abuse generative AI, covering LLM-generated and polymorphic malware, jailbreaks, and purpose-built 'dark' LLMs such as WormGPT and FraudGPT, alongside data poisoning and manipulation of models (citing 100 poisoned models found on Hugging Face). The pieces are explainers on the threat landscape rather than reports of a specific new exploitable mechanism.