Tools
52d agoHackers Use LLMs To Build Massive Botnets
Nine popular AI tools now facilitate "HalluSquatting," a technique that exploits LLM hallucinations to trick users into downloading malware. Attackers weaponize the models' inability to admit ignorance to distribute malicious links. This vulnerability turns Large Language Models into delivery vectors for botnets.
The Signal
Security practitioners must now treat AI-generated suggestions as untrusted inputs.