TECH NEWS
Cybersecurity vendors turn to multi-model, open AI to bypass guardrails | TechTarget
* Alissa Irei, Published: 29 Sep 2026 Recent high-profile security incidents highlight a growing problem across the cybersecurity industry: Commercial AI models frequently mistake legitimate security analysis for malicious activity. To prevent upstream AI safety guardrails from stalling critical security workflows, security vendors are now actively decoupling their products from single-model