Powered by the combined work of our machine learning engineers and detection team, we’ve made significant improvements to our detection accuracy.
Smarter Core Models
Our core detection models now leverage 50% more features and advanced architectures, enabling proactive identification of more than 30,000 additional attack campaigns each week, without increasing false positives.
Expanded Phishing Coverage
We’ve added 27 new detectors for internal impersonation, brand impersonation, and compromised sender attacks. These updates are already blocking over 60,000 additional phishing attempts weekly and have reduced missed detections by 30%.
Stronger Campaign Consistency
Campaign consistency has been extended beyond ATTACK and SPAM to also include SAFE and GRAYMAIL. This delivers a 4% improvement in AI Security Mailbox consistency and a 1% boost in message campaign consistency, ensuring more reliable classification across all message types.
These enhancements reflect our ongoing investment in people and AI-enabled systems, expanding protection against evolving attacks while keeping operations efficient and burden-free for security teams.