Nvidia plans to launch an open source AI agent platform

Nvidia is planning To launch an open source platform for AI clients, people familiar with the company’s plans told WIRED.

The chipmaker has been promoting the product, referred to as NemoClaw, to enterprise software companies. The platform will allow these companies to send AI agents to perform tasks for their workforce. Sources say companies will be able to access the platform regardless of whether their products run on Nvidia chips.

The move comes as Nvidia prepares for its annual developers conference in San Jose next week. Ahead of the conference, Nvidia reached out to companies including Salesforce, Cisco, Google, Adobe, and CrowdStrike to establish partnerships for the agent platform. It is unclear whether these talks have resulted in formal partnerships. Since the platform is open source, partners will likely get free early access in exchange for contributing to the project, sources say. Nvidia plans to offer security and privacy tools as part of its new open source proxy platform.

Nvidia did not respond to a request for comment. Representatives for Cisco, Google, Adobe, and CrowdStrike also did not respond to requests for comment. Salesforce did not provide a statement prior to publication.

Nvidia’s interest in agents comes at a time when people are embracing “tentacles,” or open-source AI tools that run locally on a user’s device and perform sequential tasks. Claws are often described as self-learning, as they are supposed to improve automatically over time. Earlier this year, an artificial intelligence agent known as OpenClaw— which was initially called Clawdbot, then Moltbot — took Silicon Valley by storm due to its ability to run autonomously on personal computers and complete work tasks for users. OpenAI ended up acquiring the project and hiring its creator.

OpenAI and Anthropic have made significant improvements in model reliability in recent years, but their chatbots still require hand-holding. On the other hand, AI agents or specially designed claws are designed to perform multiple steps without much human supervision.

The use of claws within enterprise environments is controversial. Some tech companies, including Meta, have done just that, WIRED previously reported I asked the employees to refrain Avoid using OpenClaw on business computers, due to the unpredictability of proxies and potential security risks. Last month, a Meta employee was overseeing safety and alignment in the company’s AI lab The story was shared publicly It’s about an AI agent who goes around her device and deletes her emails en masse.

For Nvidia, NemoClaw appears to be part of an attempt to court enterprise software companies by offering additional layers of security to its AI customers. It’s also another step up for the company Embrace open source AI modelswhich is part of a broader strategy to maintain its dominance in AI infrastructure as leading AI labs build their own custom chips. Nvidia’s software strategy to date has been largely based on its CUDA platform, a popular proprietary system that restricts developers to building software for Nvidia’s GPUs and has created a crucial “moat” for the company.

Last month’s Wall Street Journal I mentioned Nvidia is also planning to unveil a new chipset system for inferential computing at its developer conference. The system will include a chip designed by startup Groq, with which Nvidia entered into a multibillion-dollar licensing agreement late last year.

Paresh Dave contributed to this report.