ServiceNow made a series of ambitious announcements at Knowledge 2026, unveiling three closely integrated advancements that strengthen its position as the central AI control layer for large enterprises. The updates include a substantially expanded AI Control Tower, the introduction of ServiceNow Otto as a unified AI experience, and a deepened strategic partnership with NVIDIA to extend agentic AI governance from employee desktops all the way to data center infrastructure.
These launches come at a critical time as organizations scale autonomous AI agents across hybrid environments, from desktops and business applications to cloud services and large-scale AI factories. The expanded AI Control Tower now provides comprehensive discovery, real-time observation, governance, security, and ROI measurement across virtually any AI system in the enterprise. Otto embeds intelligent, multimodal AI directly into daily employee workflows, while the NVIDIA collaboration introduces Project Arc — a governed autonomous desktop agent — and integrates AI Control Tower with the NVIDIA Enterprise AI Factory validated design.
“Enterprises are under real pressure to deploy AI and show results, but there’s a major gap between adoption and accountability,”
Jon Sigler, Executive Vice President and General Manager of AI Platform at ServiceNow.
Together, these capabilities create a cohesive system that connects discovery, user experience, runtime governance, and infrastructure-level control.
Conditions Driving This Launch
Enterprises have accelerated AI adoption but struggle with fragmented visibility and control as agents run across desktops, clouds, applications, and data centers.
Traditional governance tools provide only partial coverage and cannot keep pace with dynamic, non-deterministic agent behavior.
The explosion of AI agents has created urgent needs for runtime observability, behavioral guardrails, and real-time intervention capabilities.
Organizations face increasing regulatory pressure under frameworks like the EU AI Act and NIST, demanding demonstrable oversight across all AI systems.
Runaway AI costs and security risks from uncontrolled agents are pushing boards and CISOs to demand unified accountability.
Hybrid and multi-cloud environments make centralized discovery and policy enforcement extremely difficult without a dedicated control layer.
The convergence of desktop agents and large-scale model infrastructure requires governance that spans from endpoints to data centers.
Enterprises need AI experiences that feel natural and embedded in daily work rather than isolated chatbots or separate tools.
Strategic partnerships with hyperscalers and infrastructure leaders like NVIDIA and AWS are essential to deliver governed AI at scale.
What Enterprise AI Governance Looked Like Before
Before these announcements, enterprise AI governance was fragmented, reactive, and incomplete. Organizations typically relied on disparate tools for different parts of the AI lifecycle — one solution for model governance, another for cloud security, basic logging for monitoring, and manual processes for compliance.
Visibility was limited to systems inside the ServiceNow platform, leaving shadow AI, agents running on desktops, and workloads in hyperscaler environments largely invisible. Runtime observability into how agents reasoned and made decisions was minimal, making it difficult to detect when agents deviated from expected behavior or exceeded permissions.
Security teams lacked unified kill switches or real-time enforcement across environments. Measuring ROI and controlling costs was challenging amid growing model spend. Employees interacted with AI through clunky chat interfaces that pulled them out of their natural workflows. Scaling agentic AI meant accepting significant governance, security, and compliance risks, forcing many organizations to choose between innovation speed and safety. The lack of a unified control tower created dangerous blind spots in an increasingly agent-driven enterprise.
What Enterprise AI Governance Looks Like Now
With the new releases, ServiceNow delivers a unified, end-to-end governance layer for the entire AI estate. The expanded AI Control Tower now discovers AI assets across AWS, Google Cloud, Azure, SAP, Oracle, Workday, and more, while extending to non-human identities, OT, and IoT devices. It provides continuous observation powered by the Traceloop acquisition, five new risk frameworks aligned to NIST and EU AI Act, real-time security with Veza integration, and detailed cost/ROI measurement.
ServiceNow Otto transforms the employee experience by embedding voice, vision, Dynamic Guidance, Screen Summarization, SmartDocs, and Intelligent Approvals directly into workflows. Project Arc, the new autonomous desktop agent, runs securely in NVIDIA OpenShell and is fully governed by AI Control Tower, enabling trusted autonomous work on employee desktops.
The deepened NVIDIA partnership extends governance to data centers through integration with the NVIDIA Enterprise AI Factory validated design. This creates consistent policy enforcement, observability, and control from endpoints to large-scale model infrastructure. An AI Gateway adds real-time controls for third-party models, while new benchmarks (NOWAI-Bench) help organizations measure agent performance objectively.
If your organization is scaling agentic AI across hybrid environments, visit the GAIG marketplace today to compare platforms that deliver true end-to-end governance, runtime protection, and unified AI experiences.
Our Take
AI Governance Take
ServiceNow’s trio of announcements at Knowledge 2026 represents a major leap forward in enterprise AI governance. By expanding the AI Control Tower, introducing Otto for seamless user experiences, and partnering with NVIDIA to govern agents from desktops to data centers, ServiceNow is building a true AI control layer that spans the entire enterprise stack.
These innovations address the core challenge of the agentic era: delivering the speed and autonomy organizations want while maintaining the visibility, security, and compliance they need. The integration of discovery, runtime observability, behavioral controls, and infrastructure governance creates a cohesive system that standalone tools cannot match.
Enterprises no longer have to choose between innovation and control. ServiceNow, together with NVIDIA and major cloud providers, offers a unified platform where AI can be deployed responsibly at scale. Organizations that adopt this comprehensive approach will gain competitive advantage through faster, safer, and more measurable AI outcomes.
