Intelligence needs infrastructure of trust.
Every wave of computing created a security industry in its wake. AI compresses that timeline: capability, oversight, and defense must be built simultaneously. We fund all three.
AI development
Foundation-model tooling, applied AI products, and the developer infrastructure that turns research into deployed systems.
Explore development → 02AI governance
Evaluation, audit, and compliance platforms that let enterprises adopt AI with confidence — and prove it.
Explore governance → 03AI security
Red-teaming, runtime protection, and data provenance — the defensive layer for a new attack surface.
Explore security →
The most trusted AI system wins the enterprise, the regulator, and ultimately the market. Trust is not a compliance cost — it is the product.
Capability is compounding
Model capability doubles on a cadence measured in months. Every doubling expands what software can do — and what can go wrong when it does.
Adoption is gated by trust
Enterprises and regulators don't buy capability; they buy proof — evaluation results, audit trails, security guarantees. Trust infrastructure is the bottleneck on the entire market.
The bottleneck is the opportunity
Whoever relieves the bottleneck captures a toll position on AI adoption itself. Governance and security companies sell to everyone who deploys AI — a market that is, increasingly, everyone.
Standards decide the winners
As evaluation and security practices standardize, the companies whose methods become the standard become the default. We invest before that consolidation, in the teams shaping it.