Keynote – Amit Zaveri
Summary
Zaveri framed trust as the new infrastructure for enterprise AI. He outlined ServiceNow’s AI Control Tower, which provides end‑to‑end visibility, compliance, and security for AI agents embedded in workflows.
Key components:
- Human‑in‑the‑loop design: AI augments, not replaces, human decision‑making.
- Security‑by‑design: Integration of Vesa’s non‑human identity graph to manage agent permissions.
- ROI demonstration: Early adopters see measurable efficiency gains, leading to broader rollout.
Zaveri warned that uncontrolled agentic AI can cause compliance failures and security breaches, emphasizing the need for policy‑as‑code, audit logs, and continuous monitoring. He predicted that AI adoption will accelerate once trust frameworks are embedded in enterprise platforms.
Key Takeaways
- Trust = infrastructure: Visibility, auditability, and security are prerequisites for AI adoption.
- AI Control Tower: Centralized governance for AI agents across the enterprise.
- Human‑in‑the‑loop: AI should enhance, not replace, human expertise.
- Security for agents: Identity graphs ensure agents have appropriate permissions.
- ROI‑driven rollout: Demonstrated gains accelerate broader AI deployment.