M-Trust：Trust Paradigm Evolution for Future Agentic Networks

Driven by Agentic AI, next-generation intelligent networks are shifting from being human-centric to intelligence-centric. As network architectures evolve and service paradigms transform, future networks exhibit decentralized, intelligent, and heterogeneous interconnection characteristics. While these attributes boost system performance, they introduce new security and trust challenges. Zero Trust architectures, although effective in traditional networks, encounter limitations when confronted with distributed agent interconnections, autonomous decision-making, and cross-domain interactions. Traditional centralized evaluations, static policies, and single-domain closed loops are no longer sufficient. To build trustworthy next-generation intelligent networks, this white paper proposes M-Trust, an architectural framework designed to elevate traditional Zero Trust across three dimensions：Multi-Trust、Agentic Trust and eXtensible Trust.

Download PDF（1.43MB）