CYGNVS launches AI model trained on cyber incidents
CYGNVS has launched a new AI model to support cyber readiness and incident response, positioning the tool as a purpose-built system trained on more than 20,000 real-world cyber incidents and technology outages.
The model is built to deliver incident-specific guidance during active crises, addressing a core challenge in cyber response: most organizations only experience a small number of major incidents and lack broad, hands-on exposure to complex cyber events. By contrast, the CYGNVS model draws on patterns learned from a large corpus of historical incidents to help response teams make faster, more informed decisions under pressure.
A key differentiator is the model’s training data. CYGNVS collaborated with Marsh to incorporate anonymized insights derived from Marsh’s global cyber incident experience, which spans thousands of client incidents each year. Marsh develops and vets the analytical outputs internally, ensuring no client data or identifiable information is shared, before aggregated patterns are integrated into CYGNVS’s broader training set.
According to CYGNVS, the result is an AI system that combines scale and specificity without compromising data privacy. The model is embedded within CYGNVS’s out-of-band command center, which is designed to remain accessible even when core enterprise systems are unavailable or compromised.
“Marsh has the world’s richest experience and expertise on cyber readiness and response from our 25 years in cyber insurance and our daily service to more than 15,000 clients around the world. Clients expect Marsh to continue finding new ways to make our insights available to them in moments that matter. We collaborated with CYGNVS to bring this AI model to life for our clients, enabling them to learn from and apply our experience for more successful incident readiness and response. Our approach ensures client privacy and security, while delivering proactive, context-aware AI-based guidance.” – Thomas Reagan, Global Head of Cyber at Marsh Risk.
