86% Of Enterprises Have Deployed AI Agents. Just 34% Trust Them, Boomi Study Finds.

This is less a demand shock for AI than a timing shift: enterprises are adopting agents quickly, but the gating item is now control-plane spend rather than model spend. That tends to reallocate budget toward identity, policy enforcement, auditability, and data-access tooling, which is constructive for vendors that sit in the trust stack such as CRWD, PANW, OKTA, and to a lesser extent SNOW and ZS. The first-order losers are smaller point solutions pitching autonomous agents without embedded governance; their sales cycles will lengthen and their multiple expansion depends on proving production-grade reliability, not demos.

The next 1-3 months matter more for sentiment than revenue: if software management teams start qualifying agent deployment with more security/compliance language, the market will likely haircut near-term AI uplift assumptions. Over 6-18 months, however, this can actually broaden the moat of incumbents with installed identity/data layers, because customers are more likely to buy agent workflows from platforms they already trust than from standalone startups. That favors bundled distribution at MSFT, CRM, NOW, and cloud/security vendors with policy controls baked in.

Contrarian take: the market may be overreading a trust gap that is really an integration gap. Enterprises rarely need agents to be fully autonomous; they need them to be bounded, auditable, and tied into existing systems. If that’s the case, adoption can accelerate faster than skeptics expect once vendors demonstrate measurable error rates and approval workflows. The thesis is falsified if upcoming earnings show agent features converting to >20% production penetration without elevated security spend, or if regulators force a slower compliance cycle than current procurement friction implies.