Entrust launches Agentic AI Trust Accelerator to move AI agents into production

The Authorization Bottleneck

The core problem is elementary but expensive: when an AI agent acts, who authorized it and how is that proven? Entrust's program focuses on verifiable identity for both humans and agents, cryptographic proof of action, and real-time authorization controls. This isn't about making agents smarter; it's about creating an audit trail that satisfies regulators and internal risk teams. For financial institutions and cloud providers, this infrastructure is the tax they must pay to move agents from pilot to production without assuming untenable liability.

Accelerator Structure and Entry Cost

The Accelerator is a co-development program, not a product launch. Entrust is leveraging its existing public-key infrastructure and digital certificate management to build this trust layer. Initial focus lands on four pillars: identity, authorization, cryptography, and accountability. The move signals where value will accrue in the agent economy—not just in model performance, but in the governance stack that scales with it. Entry is limited to enterprises and partners willing to co-invest development resources, a filter that will likely separate serious deployments from experimental ones.

Market Reality Check

This is a defensible market position. The $1.2 billion valuation for Entrust's identity business, per its latest round, is predicated on being the picks-and-shovels provider for digital economies. An agentic AI trust layer is a logical, high-margin extension. For the broader AI capital landscape, it's a signal that the ecosystem is maturing: the burn rate on unmonitored agents is becoming a board-level concern, and liquidity in the agent market depends on solving this. The path from pilot to production is paved with compliance costs, and Entrust is positioning itself as the toll collector.