ServiceNow and Nvidia are deepening their partnership to move enterprise AI from chaotic potential to governed control, introducing new tools to manage autonomous agents from the desktop to the data center.
ServiceNow is expanding its partnership with Nvidia to introduce Project Arc, an autonomous desktop agent designed to complete complex work, directly challenging the current workflow automation market and aiming to accelerate enterprise AI adoption with built-in governance. The move, announced at ServiceNow's Knowledge 2026 conference, signals a deeper integration between the two companies, connecting the application layer where work is done with the underlying AI infrastructure.
"Long-running, autonomous agents are rapidly changing the game for enterprise AI, and delivering them securely at scale requires governance that spans models, software and AI infrastructure," said Kari Briski, vice president of Generative AI for Enterprise at NVIDIA.
Project Arc will run within the sandboxed NVIDIA OpenShell runtime, with all actions monitored by the ServiceNow AI Control Tower, which is now integrated into the NVIDIA Enterprise AI Factory validated design for data centers. The companies also released NOWAI-Bench, an open-source framework for evaluating AI agent performance, a crucial step for establishing trust and measuring ROI.
This move positions ServiceNow (NYSE: NOW) to capture a larger share of the burgeoning enterprise AI market by providing the critical governance layer that many companies require for adoption, potentially increasing its platform's stickiness and challenging standalone automation software providers.
From Desktop to Data Center
The core of the announcement is Project Arc, an enterprise agent that lives on employee desktops and can autonomously handle multi-step tasks across different systems without pre-defined workflows. Unlike more open-ended AI assistants, every action the agent takes is contained within NVIDIA OpenShell, a secure sandbox that allows for policy-based management and auditing. ServiceNow's AI Control Tower governs the agent's behavior, logging commands, API calls, and files read, providing an audit trail that security teams can approve.
"Whether it's autonomous AI agents that can be trusted on the desktop, governance that extends to the data center, or open benchmarks that hold the entire industry accountable, this is enterprise AI that's built to last," said Joe Davis, executive vice president of AI Engineering & Delivery at ServiceNow.
The integration of the AI Control Tower with the NVIDIA Enterprise AI Factory, first previewed at GTC in March and now generally available, extends this governance to the infrastructure layer. This provides a unified management plane for organizations running large-scale AI model workloads, tracking metrics like cost, risk, and security across the entire model lifecycle.
A Standard for Measuring Agents
To address the difficulty in evaluating agentic AI systems, the companies are also releasing NOWAI-Bench as an open-source tool. The suite includes EnterpriseOps-Gym for IT and HR workflows and EVA-Bench for voice agents. Nvidia is integrating both into its NeMo Gym to promote wider industry adoption. Nvidia has already evaluated its Nemotron 3 Super model on EnterpriseOps-Gym, reporting leading performance among open-weight models.
This collaboration provides a direct challenge to other enterprise AI offerings, such as Microsoft's Copilot, by emphasizing a framework of top-down governance and security from day one. For investors, the partnership strengthens ServiceNow's position as the central AI control tower for businesses. Shares of ServiceNow, trading at a forward P/E ratio of around 55, have risen over 30% in the past year, reflecting investor confidence in its AI strategy. This expanded partnership with Nvidia, the key enabler of the current AI boom, further solidifies that narrative.
This article is for informational purposes only and does not constitute investment advice.
