AI/ML security is good.
Just not good enough.
We're big fans of AI/ML and have been working with LLMs for over three years now. In fact, we use AI/ML models every day in our global operation, and we are leveraging it across our products and services to improve automation, productivity, and throughput. But since AI/ML security models can't ever fully close the gap with AI/ML phishing, we firmly believe that a combination of approaches, including diverse intelligence from human reporting and vetting, is the only way to minimize email security risk.