ServiceNow today announced a significant expansion of its partnership with NVIDIA to extend agentic AI governance from desktops to data centers. This includes the introduction of Project Arc, a new enterprise autonomous desktop agent, secured by the NVIDIA OpenShell runtime and governed by ServiceNow AI Control Tower, that will live on employee desktops and autonomously complete complex work. ServiceNow AI Control Tower is also now included in the NVIDIA Enterprise AI Factory validated design, extending enterprise governance to large-scale model workloads. The companies are advancing enterprise AI even further with an open benchmarking standard for evaluating AI agents.
ServiceNow and NVIDIA have long partnered to bring enterprise-grade AI to some of the world’s largest organizations through open models, AI software, and agent-ready skills. From the development of specialized agents to the Apriel Nemotron open model family, the partnership has deepened at each stage of the AI era. Today’s announcement connects how work gets done with how AI runs, bringing execution, governance, and intelligence to user endpoints and AI infrastructure.
Project Arc: An enterprise autonomous desktop agent that is governed and secure from day one
ServiceNow is introducing Project Arc, which is an enterprise autonomous desktop agent that thinks, writes code, executes, and adapts when things don’t go as expected, completing complex multi-step work across enterprise tools and systems without requiring pre-built workflows.Every action the agent takes runs inside NVIDIA OpenShell, a sandboxed runtime environment that adds policy-based management so that autonomous activity stays contained, auditable, and enterprise safe. ServiceNow AI Control Tower governs the actions the agent takes, setting policies, monitoring behavior, and logging files read, commands executed, and APIs called. The result is an autonomous desktop agent that enterprise security leaders can fully audit and approve with confidence.
Powered by ServiceNow Action Fabric, which allows any AI agent to access the ServiceNow system of action, and grounded in the ServiceNow Configuration Management Database (CMDB), the agent has deep intelligence about the enterprise, pulling from workflows, systems, and operational history so that actions are informed by how work gets done. Project Arc also meets employees where they work through a desktop app, enterprise collaboration tools, or email, and is designed to provide the same governed execution, whatever the interface.
Governing large-scale AI model workloads
The ServiceNow AI Control Tower integration with the NVIDIA Enterprise AI Factory validated design was previewed at NVIDIA GTC in March and is now complete, extending enterprise governance to the infrastructure layer where large-scale AI model workloads run. For organizations managing AI at scale, this integration provides a unified governance layer for data centers. AI factory deployments governed by AI Control Tower gain continuous value, risk, and security management across the full model lifecycle, from discovery and inventory to real-time observability, compliance monitoring, and remediation.
Capabilities available to AI factory customers through AI Control Tower include expanded regulatory content packs, enterprise access maps for major cloud providers, and a runtime cost and ROI management framework that tracks productivity gains and monthly value indicators.
Measuring and advancing enterprise AI together
ServiceNow and NVIDIA are advancing NOWAI-Bench, an open benchmarking suite comprising two frameworks: EnterpriseOps-Gym, a multi-step agentic evaluation framework spanning IT service management, customer service, and HR workflows, and EVA-Bench, a voice agent evaluation framework designed for enterprise settings. Both benchmarks are generally available as open- source releases, and NVIDIA is integrating both into NeMo Gym to make them reusable and accessible for automated model evaluation across the industry.
The collaboration is already producing results. NVIDIA has evaluated NVIDIA Nemotron 3 Super on EnterpriseOps-Gym, with results demonstrating leading performance among open-weight models.
On the voice side, EVA-Bench is serving as a primary benchmark for the development of NVIDIA Nemotron VoiceChat model, with ServiceNow and NVIDIA planning to publish joint evaluation results as the model reaches availability.
Availability
Project Arc is available as an early preview. AI Control Tower integration with the NVIDIA Enterprise AI Factory validated design is generally available. NOWAI-Bench, including EnterpriseOps-Gym and EVA-Bench, is available now as an open-source release.
Related
This entry was posted on May 5, 2026 at 1:06 pm and is filed under Commentary with tags NVIDIA, ServiceNow. You can follow any responses to this entry through the RSS 2.0 feed.
You can leave a response, or trackback from your own site.
ServiceNow extends agentic AI governance from desktops to data centers with NVIDIA
ServiceNow today announced a significant expansion of its partnership with NVIDIA to extend agentic AI governance from desktops to data centers. This includes the introduction of Project Arc, a new enterprise autonomous desktop agent, secured by the NVIDIA OpenShell runtime and governed by ServiceNow AI Control Tower, that will live on employee desktops and autonomously complete complex work. ServiceNow AI Control Tower is also now included in the NVIDIA Enterprise AI Factory validated design, extending enterprise governance to large-scale model workloads. The companies are advancing enterprise AI even further with an open benchmarking standard for evaluating AI agents.
ServiceNow and NVIDIA have long partnered to bring enterprise-grade AI to some of the world’s largest organizations through open models, AI software, and agent-ready skills. From the development of specialized agents to the Apriel Nemotron open model family, the partnership has deepened at each stage of the AI era. Today’s announcement connects how work gets done with how AI runs, bringing execution, governance, and intelligence to user endpoints and AI infrastructure.
Project Arc: An enterprise autonomous desktop agent that is governed and secure from day one
ServiceNow is introducing Project Arc, which is an enterprise autonomous desktop agent that thinks, writes code, executes, and adapts when things don’t go as expected, completing complex multi-step work across enterprise tools and systems without requiring pre-built workflows.Every action the agent takes runs inside NVIDIA OpenShell, a sandboxed runtime environment that adds policy-based management so that autonomous activity stays contained, auditable, and enterprise safe. ServiceNow AI Control Tower governs the actions the agent takes, setting policies, monitoring behavior, and logging files read, commands executed, and APIs called. The result is an autonomous desktop agent that enterprise security leaders can fully audit and approve with confidence.
Powered by ServiceNow Action Fabric, which allows any AI agent to access the ServiceNow system of action, and grounded in the ServiceNow Configuration Management Database (CMDB), the agent has deep intelligence about the enterprise, pulling from workflows, systems, and operational history so that actions are informed by how work gets done. Project Arc also meets employees where they work through a desktop app, enterprise collaboration tools, or email, and is designed to provide the same governed execution, whatever the interface.
Governing large-scale AI model workloads
The ServiceNow AI Control Tower integration with the NVIDIA Enterprise AI Factory validated design was previewed at NVIDIA GTC in March and is now complete, extending enterprise governance to the infrastructure layer where large-scale AI model workloads run. For organizations managing AI at scale, this integration provides a unified governance layer for data centers. AI factory deployments governed by AI Control Tower gain continuous value, risk, and security management across the full model lifecycle, from discovery and inventory to real-time observability, compliance monitoring, and remediation.
Capabilities available to AI factory customers through AI Control Tower include expanded regulatory content packs, enterprise access maps for major cloud providers, and a runtime cost and ROI management framework that tracks productivity gains and monthly value indicators.
Measuring and advancing enterprise AI together
ServiceNow and NVIDIA are advancing NOWAI-Bench, an open benchmarking suite comprising two frameworks: EnterpriseOps-Gym, a multi-step agentic evaluation framework spanning IT service management, customer service, and HR workflows, and EVA-Bench, a voice agent evaluation framework designed for enterprise settings. Both benchmarks are generally available as open- source releases, and NVIDIA is integrating both into NeMo Gym to make them reusable and accessible for automated model evaluation across the industry.
The collaboration is already producing results. NVIDIA has evaluated NVIDIA Nemotron 3 Super on EnterpriseOps-Gym, with results demonstrating leading performance among open-weight models.
On the voice side, EVA-Bench is serving as a primary benchmark for the development of NVIDIA Nemotron VoiceChat model, with ServiceNow and NVIDIA planning to publish joint evaluation results as the model reaches availability.
Availability
Project Arc is available as an early preview. AI Control Tower integration with the NVIDIA Enterprise AI Factory validated design is generally available. NOWAI-Bench, including EnterpriseOps-Gym and EVA-Bench, is available now as an open-source release.
Share this:
Like this:
Related
This entry was posted on May 5, 2026 at 1:06 pm and is filed under Commentary with tags NVIDIA, ServiceNow. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.