Hopp til hovedinnhold
 AI-nyheter, ferdig filtrert for ledere
SISTE:

Anthropic åpner Claude Security for Enterprise-kunder • Britisk AISI: GPT-5.5 når Mythos-nivå i cybertester • OpenAI åpner GPT-5.5-Cyber kun for utvalgte forsvarere

DigiCert wants to give AI agents cryptographic identity
CIOCISOAI AgentsSecurityGovernance

DigiCert wants to give AI agents cryptographic identity

JH
Joachim Høgby
30. april 202630. april 20263 min lesingKilde:

DigiCert introduced an AI Trust architecture on April 30 to secure AI agents, models and digital content. This is not another chatbot feature. It is PKI thinking moved into the agent era: identity, integrity and audit trails for systems that act at machine speed.

What is new

DigiCert describes three parts. AI Agent Trust gives agents discovery, identity, policy, lifecycle management and auditability. AI Model Trust signs, packages and validates models from development through runtime. Content Trust uses C2PA to sign and verify digital content.

The point is verifiability. Which agent is this? What is it allowed to do? Is the model the same one that was approved? Can the origin of the content be proven?

Why leaders should care

Agent projects move quickly from demo to risk. Once agents get access to CRM, ERP, code, documents and customer data, a prompt policy in Confluence is not enough. Enterprises need machine-readable identity, policy enforcement and traceability that can survive audit.

This is a CIO and CISO issue. AI governance is no longer only about model choice and data processing agreements. It is also about certificates, signing, provenance, attestation and ownership of agent actions.

Practical consequence

Enterprise AI architecture should be treated more like secure infrastructure than an app project. Require agent identity, least privilege, signed model artifacts, runtime validation and complete logging before agents get write access in production.

DigiCert is one vendor. The direction matters more than the product: AI systems must be able to prove who they are, what they run and what they produce.

Source and date validation

Original sources: DigiCert / GlobeNewswire, "DigiCert Introduces New AI Trust Architecture for Securing AI Agents, Models, and Content", published April 30, 2026, https://www.globenewswire.com/news-release/2026/04/30/3284921/0/en/DigiCert-Introduces-New-AI-Trust-Architecture-for-Securing-AI-Agents-Models-and-Content.html. DigiCert AI Trust product page, https://www.digicert.com/solutions/ai-trust. This is within the 48-hour requirement.

📬 Likte du denne?

AI-nyheter for ledere. Kuratert av en CIO som bygger det selv. Daglig i innboksen.