DigiCert launches AI Trust Manager for agent control

DigiCert has launched AI Trust Manager, now generally available as part of the DigiCert ONE platform.

The software is designed to help organisations identify, manage and control AI agents operating within their businesses and across external systems.

AI agents are taking on tasks that include accessing sensitive data, using software tools, writing code, making decisions and completing transactions with limited human involvement. For many security teams, that has created a governance gap, as identity systems built for staff and conventional software do not easily extend to autonomous agents moving between applications, cloud environments and organisations.

DigiCert's approach centres on what it calls an AI Passport, a cryptographically signed credential that verifies an agent's identity and links it to an owner. The system also uses policy-based permissions, described as visas, to limit which systems, data and actions an agent may access, and for how long.

According to DigiCert, those credentials travel with the agent, allowing another organisation to verify its identity and permissions without using the same identity provider. A receiving organisation can still deny access in its own environment, while an automated kill switch can block prohibited actions and revoke an agent's authority if trust changes.

The launch comes as Australian businesses report a rising number of AI-related security issues. In a DigiCert survey of 1,001 IT and cybersecurity leaders, 81% of Australian organisations said they had experienced an AI-related security incident or identified an AI vulnerability in the previous year.

Daniel Sutherland, Area Vice President, ANZ, at DigiCert, linked the launch to the practical challenges companies face as AI use expands.

"AI adoption in Australia continues to accelerate, and this creates new trust challenges. Recent discussions around AI regulations in Australia have reinforced the need for organisations to understand how AI is operating within their environments, who is accountable for it, and what controls are in place when something goes wrong. That need is underscored by the 65% of Australian SMEs that cite trust in AI as a barrier to adoption," said Sutherland.

He also argued that cryptographic methods will become more important as businesses deploy growing numbers of agents and models.

"As agents and models continue to proliferate in Australian organisations, cryptographic assurance will become essential to establishing and verifying trust. This will give organisations greater confidence in the agents and models they deploy and the content those systems create, supporting safer and more responsible AI adoption," Sutherland said.

Market demand

The product enters a market where buyers are placing more emphasis on technical proof of security and governance, according to IDC. That reflects broader concern among businesses about unverifiable claims around AI safeguards.

"Enterprise AI buying has moved from trust to proof. Buyers are ranking verifiable security controls as their top priority and naming unverifiable vendor claims as their top frustration," said Grace Trinidad, Research Director, AI Security and Trust, IDC.

"Cryptographic identity, attestation and revocation for agents, models and MCP servers are the machinery that produces this proof, which in turn produces trust, and DigiCert is applying the identity discipline the internet already runs at scale to exactly that problem," Trinidad said.

DigiCert said AI Trust Manager lets enterprises discover AI agents bought from third parties or developed internally, assign each one to an accountable owner, define its permissions, and change or revoke those permissions when policy or risk conditions shift. The system can also revoke groups of agents rather than only individual instances.

Identity control

DigiCert is positioning the service as an extension of digital identity principles into AI governance. Rather than relying solely on internal badges for trusted access, it aims to create a portable identity and authorisation layer that can be checked across organisational boundaries.

That framing has drawn support from security executives working on identity and access management.

"As AI agents become more embedded in enterprise workflows, establishing clear identity and accountability will be essential to deploying them responsibly at scale," said Ajitha Choudary, Vice President of Global Security Engineering & IAM at UKG.

"We see DigiCert AI Trust as a promising approach to extending proven principles of digital identity to this new class of autonomous systems. The ability to verify an agent's identity, define what it is authorised to do, and maintain visibility into its actions could give organisations the foundation they need to adopt agentic AI with greater confidence," Choudary said.

AI Trust Manager sits within DigiCert's wider strategy to apply cryptographic verification to AI agents, models and content, with the aim of giving organisations a portable way to verify an agent's identity and authority wherever it operates.