Identity-centered security solution company Entrust Corp. today launched the Agentic AI Trust Accelerator, a co-development program that brings together enterprises and technology partners to build the identity infrastructure companies need to move autonomous AI agents from pilot projects into production.

The program targets a gap that has kept many AI agents stuck in testing. Organizations want the software to carry out real work across their systems, but they often cannot answer basic questions once it does: who authorized an agent, what it’s permitted to do and how its actions can be proven after the fact.

Entrust describes the Accelerator as a “trust plane” for autonomous AI, built on verifiable identity, real-time authorization and cryptographic proof of action that can travel across systems, partners and workflows. The approach draws on its existing work in identity verification, public-key infrastructure, digital certificates and the management of keys and secrets.

The company is pitching the program as a response to a governance problem enterprises are already reporting. An IBM Corp. study found 77% of organizations surveyed report AI adoption is outpacing their governance capabilities, with 59% of technology executives citing security and compliance as top barriers to scaling AI agents. Deloitte Touche Tohmatsu Ltd. has framed the industry’s move to a “human on the loop” model as a form of automation that still demands strong oversight.

Initial work will focus on four areas aimed at regulated, compliance-heavy environments. The first is verifiable identity for both people and agents, so every action traces back to a confirmed human principal and a unique agent. Authorization controls keep agents inside approved policies, roles and delegated authority, with a human in the loop for critical decisions. A cryptographic layer protects the keys, certificates and secrets that agents use to authenticate and transact. The fourth, accountability, provides a verifiable record of what an agent did for regulators, customers and internal risk teams.

Anudeep Parhar, Entrust’s chief operating officer for digital infrastructure, is leading the Accelerator. AI agents are advancing faster than the trust infrastructure needed to govern them, he said, and companies at every stage, from early experiments to broader rollouts, need a foundation that scales with them.

“The Agentic AI Trust Accelerator brings together customers and partners to develop practical approaches for identity, authorization, cryptographic trust and accountability that work with their existing platforms,” Parhar said.

Chief Executive Tony Ball said trust will determine how quickly companies can move agentic AI from experimentation to production.

Entrust is opening the program to a limited number of enterprise customers, financial institutions, cloud and software-as-a-service providers, systems integrators and other technology partners interested in co-developing the architectures.

Emanuel Figueroa, a senior research analyst for identity security at International Data Corp., said in the announcement that governance is shifting from a point-in-time exercise to a continuous discipline as agents take on work once handled by employees, requiring verification throughout the agent lifecycle rather than static controls.

Image: Entrust