NVIDIA and ServiceNow Unveil 'Project Arc': Autonomous AI Agents for Enterprise with Built-in Governance
Breaking: NVIDIA and ServiceNow Launch Autonomous AI Agents for Enterprise Workflows
At the ServiceNow Knowledge 2026 conference, NVIDIA founder Jensen Huang and ServiceNow CEO Bill McDermott jointly unveiled Project Arc—a self-evolving, long-running autonomous desktop agent designed for enterprise knowledge workers. The announcement marks a significant expansion of the NVIDIA-ServiceNow collaboration, combining NVIDIA's accelerated computing and open models with ServiceNow's Action Fabric and AI Control Tower to deliver secure, governed AI agents.
Unlike standalone AI assistants, Project Arc connects natively to the ServiceNow AI Platform, enabling it to access local file systems, terminals, and applications to execute complex, multistep tasks that traditional automation cannot handle. The system is built on three core requirements: open, customizable models; domain-specific skills; and security that prevents sensitive data exposure—all running on AI factories optimized for efficient tokenomics.
Enterprise-Grade Control from Day One
Project Arc leverages NVIDIA OpenShell, an open-source secure runtime for developing and deploying autonomous agents in sandboxed, policy-governed environments. ServiceNow is contributing to OpenShell to help establish a common foundation for secure enterprise-grade agent execution. Enterprises can define exactly what an agent can see, which tools it can use, and how each action is contained.
“Project Arc represents the next step in our ongoing collaboration with NVIDIA, bringing autonomous execution to the desktop,” said Jon Sigler, executive vice president and general manager of AI Platform at ServiceNow. “By combining OpenShell’s runtime layer with ServiceNow AI Control Tower, and powered by ServiceNow Action Fabric, we’re delivering the governance and security that enterprise AI requires.”
Background: From Prompts to Autonomous Action
Enterprise AI has evolved from generating content to reasoning through complex problems. Companies now face a critical question: How should AI act? Early agent systems demonstrated the potential of moving beyond simple prompts, but enterprise environments demand context, control, and consistency across real workflows. The NVIDIA-ServiceNow partnership aims to fill that gap by delivering specialized autonomous agents that are safe and easy to adopt.
The collaboration spans the full stack: NVIDIA provides accelerated computing, open models, domain-specific skills, and secure agent execution software. ServiceNow contributes enterprise workflow context via its Action Fabric and governance through its AI Control Tower. Together, they aim to bring the power of autonomous AI to developers, IT teams, and administrators without compromising security or auditability.
For more on the challenges of deploying AI at scale, see our What This Means section below.
What This Means for Enterprise AI
Project Arc signals a shift from reactive AI tools to proactive, autonomous agents that can run for long periods, adapt to changing conditions, and complete complex workflows without human intervention. For enterprises, this means potential productivity gains in areas like IT operations, software development, and administrative tasks—but only if governance and security are baked in from the start.
By using OpenShell and the ServiceNow AI Control Tower, organizations can deploy agents with granular policies, audit trails, and containment measures. This addresses one of the biggest barriers to enterprise AI adoption: the fear of losing control over sensitive data and systems. The partnership also emphasizes open models, allowing companies to customize agents for their specific needs without vendor lock-in.
However, experts caution that long-running autonomous agents require careful monitoring and robust fail-safes. “Enterprises must ensure that agents operate within clearly defined boundaries,” said an industry analyst speaking on condition of anonymity. “Project Arc’s architecture, with sandboxed execution and policy enforcement, is a step in the right direction, but real-world deployment will demand ongoing vigilance.”
Key Takeaways
- Product: Project Arc is a long-running, self-evolving autonomous desktop agent for knowledge workers.
- Partners: NVIDIA provides OpenShell runtime, accelerated computing, and open models; ServiceNow brings workflow context and governance.
- Security: Agents run in sandboxed environments with defined policies, auditability, and data exposure controls.
- Target Users: Developers, IT teams, and administrators in enterprise settings.
- Availability: Announced at ServiceNow Knowledge 2026; integration with ServiceNow Action Fabric and AI Control Tower.
For enterprises exploring autonomous AI, the NVIDIA-ServiceNow partnership offers a blueprint for combining innovation with control. The success of Project Arc will depend on how well it scales across diverse environments and whether the promised governance features hold up under real-world pressure.
Related Articles
- 10 Breakthroughs in Robotic Touch: How DAIMON Robotics Is Redefining Dexterous Manipulation
- Building Autonomous Enterprise AI Agents: A Step-by-Step Guide with NVIDIA and ServiceNow
- Founders warned: 'The business didn't need more of me, it needed a different me' – Why stepping aside can save a growing company
- How to Secure a $35 Million Series C Extension for Your Autonomous Security Firm
- Building a Resilient Validation Framework for Autonomous Coding Agents
- From Push to Precision: How the Anthbot M9 Robot Lawn Mower Transformed My Lawn Care
- The Next Frontier in Autonomous Machinery: How Robotics is Transforming Construction
- How to Assess a Vacuum Maker's Surprise Entry into Smartphones