DigiCert has announced the launch of an AI Trust architecture designed to support the security of AI systems and their outputs. The initiative is intended to address risks associated with the increasing use of artificial intelligence and to help improve the verifiability of digital systems.
As AI is adopted across more industries, traditional approaches to trust and security are being challenged. Autonomous agents can operate across enterprise environments at high speed, and AI models can introduce risks related to supply chain security, intellectual property, and system integrity. In addition, the growing volume of AI-generated content has increased attention on the need to verify authenticity.
A key issue highlighted in this context is the lack of widely adopted cryptographically verifiable mechanisms for controlling and validating AI systems. DigiCert’s approach focuses on enabling organisations to authenticate, authorise, and audit AI-related components, with the aim of improving confidence in digital interactions.
The architecture is described as a unified trust layer intended to integrate with AI agents, models, and digital content. It uses cryptographic verification across parts of the AI lifecycle to support identity management, model integrity checks, and content authenticity within a single framework.
The architecture includes several components:
- Content Trust: Uses cryptographic signatures to verify the origin of digital content and support content provenance.
- AI Agent Trust: Supports the discovery, governance, and authentication of AI agents to help ensure controlled operation.
- AI Model Trust: Provides mechanisms for secure packaging, signing, and runtime validation of AI models to help reduce the risk of tampering.
Together, these elements are intended to replace more fragmented or manual processes with an approach that supports identity verification, integrity checks, and ongoing validation across AI systems.
DigiCert positions the architecture as a response to growing needs around trust and security in AI adoption. The approach is intended to help organisations manage risks such as regulatory compliance, reputational concerns, and system integrity, while supporting the use of AI systems in enterprise environments.
As AI continues to expand across sectors, verification of content, models, and agents is increasingly viewed as an important consideration for organisations deploying these technologies.
