Nvidia is planning to launch an open-source platform for AI agents called NemoClaw, according to a report by Wired citing people familiar with the project. The platform is designed to allow companies to deploy AI agents capable of performing complex tasks across enterprise software environments.
The chipmaker has reportedly begun pitching the technology to major enterprise software vendors including Salesforce, Cisco, Google, Adobe, and CrowdStrike in an effort to establish partnerships around the platform. It remains unclear whether any formal agreements have been finalized.
According to the report, NemoClaw will allow companies to deploy AI agents that can complete tasks on behalf of employees. These agents would be able to reason through multi-step workflows and execute actions across applications, reflecting a broader shift toward agent-based AI systems in the enterprise software market.
Because the platform is expected to be open source, potential partners may gain free access to the system while contributing development resources to the project. Sources cited by Wired said early participants could receive early access in exchange for supporting the ecosystem.
Open Ecosystem and Security Features
Nvidia’s proposed platform is expected to include built-in security and privacy features aimed at addressing concerns around autonomous AI agents operating inside enterprise systems. Companies would be able to use the platform regardless of whether their products run on Nvidia hardware, potentially broadening adoption across the software ecosystem.
The initiative reflects Nvidia’s increasing focus on AI agents as organizations move beyond large language models toward systems capable of planning, reasoning, and executing tasks independently. In recent months, the company has introduced several foundational AI models designed for agent applications, including the Nemotron and Cosmos models.
Nvidia has also expanded its NeMo software platform, which helps companies manage the lifecycle of AI agents, including training, deployment, monitoring, and optimization. NemoClaw would likely extend these capabilities by providing a shared framework for building and coordinating AI agents across enterprise tools.
Growing Interest in Agent-Based AI Tools
Interest in agent-based AI has grown rapidly this year, partly driven by the rise of open-source projects that allow AI systems to operate locally on users’ machines and execute sequential tasks. One such project, OpenClaw, gained widespread attention earlier this year before being acquired by OpenAI along with its creator.
Nvidia Chief Executive Jensen Huang recently described OpenClaw as “the most important software release probably ever,” highlighting the industry’s growing focus on autonomous AI tools.
At the same time, security researchers have raised concerns about the risks associated with early AI agent frameworks, particularly when deployed in corporate environments with access to sensitive data and systems. Nvidia’s emphasis on built-in security controls suggests the company is aiming to address those risks as it targets enterprise adoption.
The reported development of NemoClaw comes as Nvidia prepares for its annual developer conference in San Jose next week, where the company is expected to unveil new software initiatives and roadmaps alongside updates to its AI hardware platforms.
