You can read about it in blog posts later. Or you can be in the room where the next layer of AI infrastructure is actually being built.

At Confidential Computing Summit, the teams who are shaping verifiably secure AI systems—from the silicon to the agent layer—aren’t just speaking. They’re showing you how they do it.

• Mark Russinovich – CTO of Microsoft Azure
  How Azure is securing LLM workloads in production—without slowing them down.
• Jason Clinton – CISO, Anthropic
  Why trust in AI starts with runtime guarantees, not marketing slides.
• João Moura – CEO, CrewAI
  Agentic orchestration without chaos. Yes, it’s possible.
• LangChain – Core contributors
  Turning LLM playgrounds into production pipelines.
• Atindriyo Sanyal – CTO, Galileo
  Observability and safety from the prompt to the policy.

Meta: Building Trusted Execution for LLMs at Planet Scale

Speakers: Technical leads from Meta’s AI Infra team

Focus: Confidential compute architectures powering open-weight models and safety-first inference pipelines

Key Takeaways:

• How Meta uses verifiable compute for secure AI deployment at scale
• Lessons from operating large-scale open model inference across environments
• Why trusted execution matters—even when the model is open

 

Microsoft Azure: Securing GenAI from Prompt to Production

Speaker: Mark Russinovich, CTO of Microsoft Azure
Focus: Azure’s confidential AI strategy, including runtime attestation and policy-based controls

Key Takeaways:

• Confidential computing as the foundation for scalable GenAI services
• How Microsoft embeds governance without slowing delivery
• What runtime attestation looks like in practice 

 

CrewAI: Orchestrating Agent Workflows Without Losing Control

Speaker: João Moura, CEO
Focus: How enterprises can deploy autonomous agents without chaos or compliance risk

Key Takeaways:

• The architecture behind CrewAI’s enterprise-ready orchestration
• Policy enforcement and fallback logic for multi-agent systems
• Best practices for deploying agents that behave

 

Beyond Explainability: Why Provable Trust Is the New Foundation for Enterprise AI

Panelists: Mark Russinovich (CTO, Microsoft), James Kaplan (CTO, McKinsey), Nelly Porter (Google), Aaron Fulkerson (CEO, OPAQUE)

Moderator: Don Clark, The New York Times
Focus: Why interpretability alone isn’t enough—and how provable trust redefines enterprise AI infrastructure

Key Takeaways:

• Interpretability breaks down in distributed, agentic systems
• Confidential computing and attestation are enabling verifiable pipelines
• Provenance and enforcement of policy are required to scale AI in compliance-sensitive environments

Securing the AI Horizon: Enterprise-Grade Frameworks for Trustworthy AI Innovation

Panelists: Jason Clinton (CISO, Anthropic), Ion Stoica (Executive Chairman, Databricks), Teresa Tung (Accenture), Raluca Ada Popa (Google DeepMind), Daniel Rohrer (NVIDIA)
Moderator: Frederic Lardinois, The New Stack
Focus: Implementing open, multi-cloud, confidential AI with security and compliance by design
Key Takeaways:

• Building enterprise AI with verifiable cryptographic foundations
• AI Infrastructure Is the New Competitive Moat
• And the ones building it aren’t waiting.
• Confidential Computing Summit brings together the leaders who’ve made verifiability, privacy, and performance non-negotiable—because they have to. This isn’t theoretical. It’s operational.

And the ones building it aren’t waiting.

• Don Clark – The New York Times
• Jonathan Vanian – CNBCAnna
• Anna Tong – Reuters
• Megan Morrone – Axios
• Frederic Lardinois – The New Stack
• Iain Thomson – The Register
• Victor Dey – Fast Company
• Adrian Bridgewater – Forbes

Discount codes applied at checkout.

Reserve Your Spot

 

Need an analyst, academic, or group pass?
Contact tickets@confidentialcomputingsummit.com

 

Organized by OPAQUE

© 2025 OPAQUE All rights reserved.  | OPAQUE, The Confidential AI CompanyTM  |  Privacy Policy  |.  Terms of Service