Panel Discussion: Proof in Production: Customer Lessons from Confidential AI
summary
Enterprises in regulated sectors deploy confidential computing to get verifiable artifacts instead of promises: runtime attestation shows where code ran, cryptographic signatures produce cryptographically signed audit trails, and an AI control tower tracks agent deployments and risk scores. EU rules require runtime proof by Jan 2027, auditors run AI to check every decision, and collaborations with vendors (Anthropic, Opaque, ServiceNow) are producing prototypes. Immediate priorities: structured governance, strict data classification, small adversarial red teams, and low‑risk pilot pipelines that prove traceability end to end.
AI-written summary · about
chapters (9)
- 0:00Panelist introductions
- 1:35Trust limits in AI projects
- 5:00Runtime proof for regulations
- 9:33Data protection in fraud/identity
- 13:07Healthcare data governance
- 15:32Securing ServiceNow agents
- 18:03Agentic workflows in finance
- 19:49AI deployment milestones
- 24:04Closing remarks
related talks
19mHow Can We Have Private Access to Intelligence?
36mSecurity for AI and Multi-Party Collaboration with Confidential Computing (Oh, and...
37mConfidential AI: Enabling Trustworthy Ecosystems in the Age of Agentic Intelligence - Session
13mProvably Secure AI for Health Data – Powered by Confidential Computing
26mObjection! AI Security Mistakes on Trial With Kubeflow and Confidential Computing
31mMaking Trustworthy AI a Business Advantage
23mConfidential computing: from niche to mainstream
31mSecuring the Future: Navigating AI Vulnerabilities and Evolving Security Practices