Cyber insurance doesn't have a shortage of data. AI is about to prove it.

August 25, 2026

Photo by Claudio Schwarz on Unsplash

Carriers already make underwriting decisions from fragmented evidence: applications, external scans, vendor scores, interviews and historical claims.

Meanwhile, the underlying risk can change faster than the underwriting process.

Now AI changes the equation again. AI-powered security tools are making it possible to identify vulnerabilities at a speed and scale that was previously impossible. Anthropic recently reported finding more than 10,000 high- and critical-severity vulnerabilities through its Project Glasswing initiative using Claude Mythos Preview.

So what happens when the problem is no longer finding cybersecurity evidence, but understanding what all that evidence means?

For insurers, more evidence doesn't necessarily mean better underwriting.

As AI dramatically increases the amount of cybersecurity evidence available, carriers still need to determine which evidence matters, how much confidence to place in it, and what underwriting decision should follow.

The opportunity is to turn fragmented evidence about vulnerabilities, controls and emerging threats into a coherent, decision-ready view of risk, translating complex cybersecurity signals into consistent, explainable underwriting intelligence.

That is the problem we are solving at Veyrdune.

As cybersecurity becomes increasingly AI-driven, we believe the competitive advantage for carriers won't come from having more data.

It will come from knowing which evidence matters, and being able to act on it faster.