A2A and MCP security - engineering trust, identity and interoperability for agentic AI

Moderator: Philip Laider, Director, SLT Partners

12:15 - 12:30
Building trust in agentic AI: How agents collaborate securely across domains

• Networks now coordinate intelligence—across vehicles, robots, sensors, and the network itself. The key challenge: enabling AI systems to act together across provider and domain boundaries.
• The industry's biggest gap isn't AI ambition—it's trust infrastructure. Over half of telecom AI projects never reach production due to missing foundations for trusted, cross-boundary intelligence.
• Multi-agent systems need standardized interaction protocols like A2A and MCP for discovery, negotiation, and collaboration across organizational lines. Equally critical is domain-specific network AI that unifies heterogeneous elements without compromising sovereignty or compliance.
• The TM Forum AI & Data Mission is chartering the industry response. This session presents a scalable architecture for trusted AI—covering system identity, shared intelligence across providers, and the path from isolated PoCs to interoperable systems at scale.
• The session outlines the blueprint's building blocks: governance standards, domain-specific AI as sovereign foundations, and cloud-native security architectures built for multi-agent governance—engineering trust into AI-native networks from the ground up.

Kaniz Mahdi, Director Technology, AWS Industries

12:30 - 12:45
Securing the Agentic NOC: Engineering Trust, Identity and Governance for Multi-Agent Telecom Operations 

• How A2A and MCP enable secure agent-to-agent collaboration across domains and partners 
• How identity, access control, and trust are designed into AI-native, agentic architectures 
• A production-ready security blueprint for interoperable, multi-agent systems 

Jay Naillon, Vice President and Global Head of Solutions, CMET, Sutherland Global 
Mike Witkowski, Vice President Network Engineering & Operations, T-Mobile USA 
Joji Philip, Global Telecom GTM Head, Snowflake 

12:45 - 13:00
Case study: Secure agent-to-agent collaboration: Designing AI with trust and identity in mind

• Strategies for embedding identity verification and access control in agent interactions
• Protocols to maintain data integrity, confidentiality and accountability
• Practical examples of secure collaboration driving business outcomes

Ameya Talwalkar, Founder and CEO, Cequence 

Timings are provisional only and may be subject to change.