en.Wedoany.com Reported - Patronus AI announced the completion of a $50 million Series B funding round led by Greenfield Partners, and the launch of its Digital World Models, a new class of large-scale simulation environments designed to help AI systems train, evaluate, and optimize within complex digital workflows. The round included participation from existing investors Notable Capital, Lightspeed Venture Partners, Datadog, Samsung, Factorial Capital, Gokul Rajaram, as well as several leading AI and software enterprise executives.
Since its founding less than three years ago, Patronus AI has become a provider of evaluation, simulation infrastructure, and reliability testing for cutting-edge AI systems. The company currently collaborates with most of the world's top frontier AI labs and hyperscale cloud providers. Its revenue has grown more than 15-fold over the past year, reflecting rising demand for infrastructure that helps organizations train, evaluate, and deploy increasingly autonomous AI systems. The new funding brings Patronus AI's total capital raised to $70 million.
Patronus AI was co-founded by AI researchers and engineers from institutions including Meta AI, Amazon AGI, and Google, among them former Meta AI researchers Anand Kannappan and Rebecca Qian. The team's expertise spans large language model (LLM) evaluation, AI alignment, fairness, and embodied agents.
The first generation of generative AI was built on static internet text and benchmark leaderboards. But as agents enter longer, more complex workflows, the limitations of this approach are becoming increasingly apparent. An agent handling customer escalations, operating enterprise software, researching across thousands of documents, or debugging production infrastructure cannot be trained solely through benchmark memorization. These systems require dynamic environments similar to the digital worlds in which they will actually operate. Patronus AI is building what it calls Digital World Models—language-diffusion world models designed to scale the generation of simulation data for training and evaluating AI agent behavior in complex digital workflows.
The simulation infrastructure built by the company allows AI systems to train on realistic software, research, communication, and enterprise workflows. The goal is not to optimize narrow benchmark performance, but to produce agents capable of reliable operation in ambiguous, long-horizon tasks. Patronus AI believes simulation will become one of the defining infrastructure layers of the AI era. This approach aims to address the problem of scalable oversight in AI. Anand Kannappan, co-founder and CEO of Patronus AI, stated: "Static evaluations only tell you whether a model can answer a narrow question in a controlled environment. They don't tell you whether an agent can handle ambiguity, recover from failure, or operate reliably in long, unpredictable workflows. This requires environments where systems can practice, adapt, and accumulate experience over time."
With the new capital, Patronus AI plans to expand its research organization, grow its engineering team, and invest in the compute and infrastructure needed to train and run Digital World Models at scale. Itay Inbar, Partner at Greenfield Partners, noted that the future of AI will depend on systems capable of reliably learning and operating in complex environments, and that simulation is becoming key to achieving this.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
