``
Skip to main content
Back to TM Forum

Securing the agentic telco: Cyber resilience in an AI-run world

October 7, 2026
12:00-13:00
Innovate Stage
Trustworthy AI & Data

Chair: Maty Bohacek, AI Researcher, Stanford University 

Fireside with AWS & TM Forum

Kaniz Mahdi, Director Technology, AWS Industries
Guy Lupo, EVP Trustworthy AI & Data, TM Forum

The rise of enterprise SuperAgents: Secure AI at scale 

  • Scale AI beyond experimentation: Build autonomous AI agents that can execute complex enterprise tasks and drive real business outcomes 
  • Innovate securely and responsibly: Combine policy-driven governance, secure AI development frameworks and strong controls to protect enterprise data, employees and customers. 
  • Accelerate transformation and reduce costs: Use AI to modernize software delivery, speed up enterprise transformation and significantly reduce AI operating costs. 
Tom Pitassi, Chief AI Architect , Chief Data Office, AT&T 
Praveen Vemulapalli, Director, Data & Gen AI Architecture, AT&T 

Beyond oversight: Zero-trust, zero-overhead compliance for Autonomous Networks
  • Embed the control scope in every autonomous action: Carry core assurance expectations from NIST SP 800-53, ISO/IEC 27000, PCI DSS, GDPR/privacy, and NIST AI RMF directly into autonomous network orchestration
  • Use standards to shape autonomic decision logic: Convert security, privacy, and compliance controls into practical guardrails for heuristic development, real-time state verification, escalation, and rollback
  • Continuously test failure modes with Risk-Based Continuous Adversarial Red Teaming: Use multiple (air-gapped) models and training strategies to challenge autonomous decisions, expose agentic logic failures, and strengthen runtime assurance before deployment.
Michael Fountain, Principle Technology Security Director, AT&T

Building trust in agentic AI with observability 
  • End-to-end observability across agent decisions, actions and business outcomes for complete operational visibility 
  • Build trust and confidence with explainability, governance, policy guardrails and continuous monitoring of AI agents 
  • Scale Agentic AI with confidence by measuring performance, detecting anomalies and ensuring reliable, compliant operations. 
Sruthi Nair, Director, Voice Core Engineering and Automation, Boost Mobile
Share session